diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/generation_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/model.safetensors.index.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..5482b26c0c0ef0ead69b4f61b1d0cb8b90f3e90e --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/model.safetensors.index.json @@ -0,0 +1,722 @@ +{ + "metadata": { + "total_parameters": 4826771968, + "total_size": 9653543936 + }, + "weight_map": { + "lm_head.weight": "model-00001-of-00002.safetensors", + "model.language_model.embed_tokens.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.merger.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.merger.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.patch_embed.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.patch_embed.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.pos_embed.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/processor_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/special_tokens_map.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/tokenizer_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_30/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/added_tokens.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/chat_template.jinja b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fa428720af4bb46e922d1d80535378a7a55db8dd --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/config.json @@ -0,0 +1,68 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "float32", + "eos_token_id": 151645, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "float32", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9728, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_theta": 5000000, + "tie_word_embeddings": true, + "use_cache": true, + "vocab_size": 151936, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_theta": 5000000, + "rope_type": "default" + } + }, + "tie_word_embeddings": true, + "transformers_version": "5.0.0.dev0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 5, + 11, + 17 + ], + "depth": 24, + "dtype": "float32", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/generation_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/model.safetensors.index.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..c27303433ba7916b58abf9596b553239720665ab --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/model.safetensors.index.json @@ -0,0 +1,722 @@ +{ + "metadata": { + "total_parameters": 4826771968, + "total_size": 9653543936 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.language_model.embed_tokens.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.patch_embed.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.patch_embed.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.pos_embed.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/processor_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/special_tokens_map.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/tokenizer_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_40/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/added_tokens.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/chat_template.jinja b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fa428720af4bb46e922d1d80535378a7a55db8dd --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/config.json @@ -0,0 +1,68 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "float32", + "eos_token_id": 151645, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "float32", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9728, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_theta": 5000000, + "tie_word_embeddings": true, + "use_cache": true, + "vocab_size": 151936, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_theta": 5000000, + "rope_type": "default" + } + }, + "tie_word_embeddings": true, + "transformers_version": "5.0.0.dev0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 5, + 11, + 17 + ], + "depth": 24, + "dtype": "float32", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/generation_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/model.safetensors.index.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..7ea24e01346ef93573413b60dff96339948f8e3c --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/model.safetensors.index.json @@ -0,0 +1,722 @@ +{ + "metadata": { + "total_parameters": 4826771968, + "total_size": 9653543936 + }, + "weight_map": { + "lm_head.weight": "model-00001-of-00002.safetensors", + "model.language_model.embed_tokens.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.merger.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.merger.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.patch_embed.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.patch_embed.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.pos_embed.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/processor_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/special_tokens_map.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/tokenizer_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_60/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/added_tokens.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/chat_template.jinja b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fa428720af4bb46e922d1d80535378a7a55db8dd --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/config.json @@ -0,0 +1,68 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "float32", + "eos_token_id": 151645, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "float32", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9728, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_theta": 5000000, + "tie_word_embeddings": true, + "use_cache": true, + "vocab_size": 151936, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_theta": 5000000, + "rope_type": "default" + } + }, + "tie_word_embeddings": true, + "transformers_version": "5.0.0.dev0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 5, + 11, + 17 + ], + "depth": 24, + "dtype": "float32", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/generation_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/model.safetensors.index.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..58e8850ffd4c8c428f208b6bea41806a1581cdf6 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/model.safetensors.index.json @@ -0,0 +1,722 @@ +{ + "metadata": { + "total_parameters": 4826771968, + "total_size": 9653543936 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00003.safetensors", + "model.language_model.embed_tokens.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00003-of-00003.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00003-of-00003.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00003-of-00003.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00003-of-00003.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00003-of-00003.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00003-of-00003.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00003-of-00003.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00003-of-00003.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00003-of-00003.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00003-of-00003.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00003-of-00003.safetensors", + "model.language_model.norm.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00001-of-00003.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00001-of-00003.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00001-of-00003.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00002-of-00003.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00001-of-00003.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00001-of-00003.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00002-of-00003.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00001-of-00003.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00003-of-00003.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00001-of-00003.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00001-of-00003.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00001-of-00003.safetensors", + "model.visual.merger.norm.bias": "model-00001-of-00003.safetensors", + "model.visual.merger.norm.weight": "model-00002-of-00003.safetensors", + "model.visual.patch_embed.proj.bias": "model-00001-of-00003.safetensors", + "model.visual.patch_embed.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.pos_embed.weight": "model-00001-of-00003.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/processor_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/special_tokens_map.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/tokenizer_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_70/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/added_tokens.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/chat_template.jinja b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fa428720af4bb46e922d1d80535378a7a55db8dd --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/config.json @@ -0,0 +1,68 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "float32", + "eos_token_id": 151645, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "float32", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9728, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_theta": 5000000, + "tie_word_embeddings": true, + "use_cache": true, + "vocab_size": 151936, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_theta": 5000000, + "rope_type": "default" + } + }, + "tie_word_embeddings": true, + "transformers_version": "5.0.0.dev0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 5, + 11, + 17 + ], + "depth": 24, + "dtype": "float32", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/generation_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/model.safetensors.index.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..92093c9c190729b892e3c89ffa80462af69147ac --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/model.safetensors.index.json @@ -0,0 +1,722 @@ +{ + "metadata": { + "total_parameters": 4826771968, + "total_size": 9653543936 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.language_model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.merger.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.patch_embed.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.patch_embed.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.pos_embed.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/processor_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/special_tokens_map.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/tokenizer_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_80/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/added_tokens.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/chat_template.jinja b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fa428720af4bb46e922d1d80535378a7a55db8dd --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/config.json @@ -0,0 +1,68 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "float32", + "eos_token_id": 151645, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "float32", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9728, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_theta": 5000000, + "tie_word_embeddings": true, + "use_cache": true, + "vocab_size": 151936, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_theta": 5000000, + "rope_type": "default" + } + }, + "tie_word_embeddings": true, + "transformers_version": "5.0.0.dev0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 5, + 11, + 17 + ], + "depth": 24, + "dtype": "float32", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/generation_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/model.safetensors.index.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..476f0b8dc9227e12cd5a26b30de76c46c67d80fc --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/model.safetensors.index.json @@ -0,0 +1,722 @@ +{ + "metadata": { + "total_parameters": 4826771968, + "total_size": 9653543936 + }, + "weight_map": { + "lm_head.weight": "model-00001-of-00002.safetensors", + "model.language_model.embed_tokens.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.merger.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.merger.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.patch_embed.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.patch_embed.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.pos_embed.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/processor_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/special_tokens_map.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/tokenizer_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_90/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/added_tokens.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/chat_template.jinja b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fa428720af4bb46e922d1d80535378a7a55db8dd --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/config.json @@ -0,0 +1,68 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "float32", + "eos_token_id": 151645, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "float32", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9728, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_theta": 5000000, + "tie_word_embeddings": true, + "use_cache": true, + "vocab_size": 151936, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_theta": 5000000, + "rope_type": "default" + } + }, + "tie_word_embeddings": true, + "transformers_version": "5.0.0.dev0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 5, + 11, + 17 + ], + "depth": 24, + "dtype": "float32", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/generation_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/model.safetensors.index.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..82ab55bf3d934f728a7cb3c6c781f6a4816193eb --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/model.safetensors.index.json @@ -0,0 +1,722 @@ +{ + "metadata": { + "total_parameters": 4826771968, + "total_size": 9653543936 + }, + "weight_map": { + "lm_head.weight": "model-00001-of-00002.safetensors", + "model.language_model.embed_tokens.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00002.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00002-of-00002.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00002-of-00002.safetensors", + "model.visual.merger.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.patch_embed.proj.bias": "model-00002-of-00002.safetensors", + "model.visual.patch_embed.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.pos_embed.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/processor_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/special_tokens_map.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/tokenizer_config.json b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_4b_ntr_multimath_gemini_50k_global_step_100_multimath_rl_50k/global_step_97/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/args.json b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/args.json new file mode 100644 index 0000000000000000000000000000000000000000..c989f87bdfe929c0b90e46b8d8e59d4847491c96 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/args.json @@ -0,0 +1,386 @@ +{ + "output_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442", + "overwrite_output_dir": false, + "do_train": false, + "do_eval": false, + "do_predict": false, + "eval_strategy": "no", + "prediction_loss_only": false, + "per_device_train_batch_size": 4, + "per_device_eval_batch_size": 4, + "per_gpu_train_batch_size": null, + "per_gpu_eval_batch_size": null, + "gradient_accumulation_steps": 4, + "eval_accumulation_steps": null, + "eval_delay": 0, + "torch_empty_cache_steps": null, + "learning_rate": 1e-05, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "max_grad_norm": 1.0, + "num_train_epochs": 1.0, + "max_steps": -1, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_ratio": 0.03, + "warmup_steps": 0, + "log_level": "passive", + "log_level_replica": "warning", + "log_on_each_node": true, + "logging_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/runs", + "logging_strategy": "steps", + "logging_first_step": true, + "logging_steps": 5, + "logging_nan_inf_filter": true, + "save_strategy": "epoch", + "save_steps": 500, + "save_total_limit": 2, + "save_safetensors": true, + "save_on_each_node": false, + "save_only_model": false, + "restore_callback_states_from_checkpoint": false, + "no_cuda": false, + "use_cpu": false, + "use_mps_device": false, + "seed": 42, + "data_seed": 42, + "jit_mode_eval": false, + "bf16": true, + "fp16": false, + "fp16_opt_level": "O1", + "half_precision_backend": "auto", + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": null, + "local_rank": 0, + "ddp_backend": null, + "tpu_num_cores": null, + "tpu_metrics_debug": false, + "debug": null, + "dataloader_drop_last": false, + "eval_steps": null, + "dataloader_num_workers": 24, + "dataloader_prefetch_factor": 2, + "past_index": -1, + "run_name": "qwen3vl_4b_sft_timelens100k_1epoch", + "disable_tqdm": null, + "remove_unused_columns": true, + "label_names": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "fsdp": [], + "fsdp_min_num_params": 0, + "fsdp_config": null, + "fsdp_transformer_layer_cls_to_wrap": null, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "deepspeed": { + "fp16": { + "enabled": "auto", + "loss_scale": 0, + "loss_scale_window": 1000, + "initial_scale_power": 16, + "hysteresis": 2, + "min_loss_scale": 1 + }, + "bf16": { + "enabled": "auto" + }, + "zero_optimization": { + "stage": 2, + "offload_optimizer": { + "device": "none", + "pin_memory": true + }, + "allgather_partitions": true, + "allgather_bucket_size": 200000000.0, + "overlap_comm": false, + "reduce_scatter": true, + "reduce_bucket_size": 200000000.0, + "contiguous_gradients": true + }, + "gradient_accumulation_steps": "auto", + "gradient_clipping": "auto", + "steps_per_print": 2000, + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "wall_clock_breakdown": false + }, + "label_smoothing_factor": 0.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "adafactor": false, + "group_by_length": false, + "length_column_name": "length", + "report_to": [ + "wandb" + ], + "project": "huggingface", + "trackio_space_id": "trackio", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": false, + "skip_memory_metrics": true, + "use_legacy_prediction_loop": false, + "push_to_hub": false, + "resume_from_checkpoint": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_token": null, + "hub_private_repo": null, + "hub_always_push": false, + "hub_revision": null, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "include_inputs_for_metrics": false, + "include_for_metrics": [], + "eval_do_concat_batches": true, + "fp16_backend": "auto", + "push_to_hub_model_id": null, + "push_to_hub_organization": null, + "push_to_hub_token": null, + "mp_parameters": "", + "auto_find_batch_size": false, + "full_determinism": false, + "torchdynamo": null, + "ray_scope": "last", + "ddp_timeout": 18000000, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "include_tokens_per_second": false, + "include_num_input_tokens_seen": false, + "neftune_noise_alpha": null, + "optim_target_modules": null, + "batch_eval_metrics": false, + "eval_on_start": false, + "use_liger_kernel": false, + "liger_kernel_config": null, + "eval_use_gather_object": false, + "average_tokens_across_devices": true, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": null, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "metric": [ + "temporal_grounding" + ], + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "Qwen/Qwen3-VL-4B-Instruct", + "model_type": "qwen3_vl", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": "flash_attn", + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_vl", + "system": null, + "max_length": 262144, + "truncation_strategy": "delete", + "max_pixels": null, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_free": false, + "padding_side": "right", + "loss_scale": "default", + "sequence_parallel_size": 4, + "response_prefix": null, + "template_backend": "swift", + "dataset": [ + "../data/timelens-100k/train/train.jsonl" + ], + "val_dataset": [], + "cached_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 1, + "load_from_cache_file": true, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "ckpt_dir": null, + "lora_modules": [], + "train_type": "full", + "adapters": [], + "external_plugins": [], + "model_kwargs": {}, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "custom_register_path": [], + "use_hf": true, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [ + "model.visual.merger", + "model.visual.deepstack_merger_list" + ], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": false, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": null, + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_lark_webhook_url": null, + "swanlab_lark_secret": null, + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "early_stop_interval": null, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3-VL-4B-Instruct", + "model_info": "ModelInfo(model_type='qwen3_vl', model_dir='/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling={'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default'}, is_moe_model=False, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_vl', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct', hf_model_id='Qwen/Qwen3-VL-2B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking', hf_model_id='Qwen/Qwen3-VL-2B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct', hf_model_id='Qwen/Qwen3-VL-4B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking', hf_model_id='Qwen/Qwen3-VL-4B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct', hf_model_id='Qwen/Qwen3-VL-8B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking', hf_model_id='Qwen/Qwen3-VL-8B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct', hf_model_id='Qwen/Qwen3-VL-32B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking', hf_model_id='Qwen/Qwen3-VL-32B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='qwen3_vl', get_function=, model_arch=MultiModelKeys(arch_name='qwen3_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model'], aligner=['model.visual.merger', 'model.visual.deepstack_merger_list'], vision_tower=['model.visual'], generator=[]), architectures=['Qwen3VLForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, is_reranker=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])", + "model_dir": "/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17", + "_val_dataset_exists": [], + "hub": "", + "evaluation_strategy": "epoch", + "training_args": "Seq2SeqTrainingArguments(output_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442', overwrite_output_dir=False, do_train=False, do_eval=False, do_predict=False, eval_strategy=, prediction_loss_only=False, per_device_train_batch_size=4, per_device_eval_batch_size=4, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=4, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=1e-05, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=1.0, max_steps=-1, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_ratio=0.03, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/runs', logging_strategy=, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=, save_steps=500, save_total_limit=2, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, jit_mode_eval=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=None, dataloader_num_workers=24, dataloader_prefetch_factor=2, past_index=-1, run_name='qwen3vl_4b_sft_timelens100k_1epoch', disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, deepspeed={'fp16': {'enabled': 'auto', 'loss_scale': 0, 'loss_scale_window': 1000, 'initial_scale_power': 16, 'hysteresis': 2, 'min_loss_scale': 1}, 'bf16': {'enabled': 'auto'}, 'zero_optimization': {'stage': 2, 'offload_optimizer': {'device': 'none', 'pin_memory': True}, 'allgather_partitions': True, 'allgather_bucket_size': 200000000.0, 'overlap_comm': False, 'reduce_scatter': True, 'reduce_bucket_size': 200000000.0, 'contiguous_gradients': True}, 'gradient_accumulation_steps': 'auto', 'gradient_clipping': 'auto', 'steps_per_print': 2000, 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'wall_clock_breakdown': False}, label_smoothing_factor=0.0, optim=, optim_args=None, adafactor=False, group_by_length=False, length_column_name='length', report_to=['wandb'], project='huggingface', trackio_space_id='trackio', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint=None, hub_model_id=None, hub_strategy=, hub_token=None, hub_private_repo=None, hub_always_push=False, hub_revision=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=18000000, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, include_tokens_per_second=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, liger_kernel_config=None, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=True, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, check_model=True, acc_strategy='token', train_dataloader_shuffle=True, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, metric=['temporal_grounding'], eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, use_flash_ckpt=False, sft_alpha=0, chord_sft_dataset=[], chord_sft_per_device_train_batch_size=None, chord_enable_phi_function=False, chord_mu_warmup_steps=None, chord_mu_decay_steps=None, chord_mu_peak=None, chord_mu_valley=None, train_type='full', local_repo_path=None, galore_config=None, padding_side='right', padding_free=False, task_type='causal_lm', problem_type=None)" +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/added_tokens.json b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/args.json b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/args.json new file mode 100644 index 0000000000000000000000000000000000000000..c989f87bdfe929c0b90e46b8d8e59d4847491c96 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/args.json @@ -0,0 +1,386 @@ +{ + "output_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442", + "overwrite_output_dir": false, + "do_train": false, + "do_eval": false, + "do_predict": false, + "eval_strategy": "no", + "prediction_loss_only": false, + "per_device_train_batch_size": 4, + "per_device_eval_batch_size": 4, + "per_gpu_train_batch_size": null, + "per_gpu_eval_batch_size": null, + "gradient_accumulation_steps": 4, + "eval_accumulation_steps": null, + "eval_delay": 0, + "torch_empty_cache_steps": null, + "learning_rate": 1e-05, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "max_grad_norm": 1.0, + "num_train_epochs": 1.0, + "max_steps": -1, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_ratio": 0.03, + "warmup_steps": 0, + "log_level": "passive", + "log_level_replica": "warning", + "log_on_each_node": true, + "logging_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/runs", + "logging_strategy": "steps", + "logging_first_step": true, + "logging_steps": 5, + "logging_nan_inf_filter": true, + "save_strategy": "epoch", + "save_steps": 500, + "save_total_limit": 2, + "save_safetensors": true, + "save_on_each_node": false, + "save_only_model": false, + "restore_callback_states_from_checkpoint": false, + "no_cuda": false, + "use_cpu": false, + "use_mps_device": false, + "seed": 42, + "data_seed": 42, + "jit_mode_eval": false, + "bf16": true, + "fp16": false, + "fp16_opt_level": "O1", + "half_precision_backend": "auto", + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": null, + "local_rank": 0, + "ddp_backend": null, + "tpu_num_cores": null, + "tpu_metrics_debug": false, + "debug": null, + "dataloader_drop_last": false, + "eval_steps": null, + "dataloader_num_workers": 24, + "dataloader_prefetch_factor": 2, + "past_index": -1, + "run_name": "qwen3vl_4b_sft_timelens100k_1epoch", + "disable_tqdm": null, + "remove_unused_columns": true, + "label_names": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "fsdp": [], + "fsdp_min_num_params": 0, + "fsdp_config": null, + "fsdp_transformer_layer_cls_to_wrap": null, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "deepspeed": { + "fp16": { + "enabled": "auto", + "loss_scale": 0, + "loss_scale_window": 1000, + "initial_scale_power": 16, + "hysteresis": 2, + "min_loss_scale": 1 + }, + "bf16": { + "enabled": "auto" + }, + "zero_optimization": { + "stage": 2, + "offload_optimizer": { + "device": "none", + "pin_memory": true + }, + "allgather_partitions": true, + "allgather_bucket_size": 200000000.0, + "overlap_comm": false, + "reduce_scatter": true, + "reduce_bucket_size": 200000000.0, + "contiguous_gradients": true + }, + "gradient_accumulation_steps": "auto", + "gradient_clipping": "auto", + "steps_per_print": 2000, + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "wall_clock_breakdown": false + }, + "label_smoothing_factor": 0.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "adafactor": false, + "group_by_length": false, + "length_column_name": "length", + "report_to": [ + "wandb" + ], + "project": "huggingface", + "trackio_space_id": "trackio", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": false, + "skip_memory_metrics": true, + "use_legacy_prediction_loop": false, + "push_to_hub": false, + "resume_from_checkpoint": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_token": null, + "hub_private_repo": null, + "hub_always_push": false, + "hub_revision": null, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "include_inputs_for_metrics": false, + "include_for_metrics": [], + "eval_do_concat_batches": true, + "fp16_backend": "auto", + "push_to_hub_model_id": null, + "push_to_hub_organization": null, + "push_to_hub_token": null, + "mp_parameters": "", + "auto_find_batch_size": false, + "full_determinism": false, + "torchdynamo": null, + "ray_scope": "last", + "ddp_timeout": 18000000, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "include_tokens_per_second": false, + "include_num_input_tokens_seen": false, + "neftune_noise_alpha": null, + "optim_target_modules": null, + "batch_eval_metrics": false, + "eval_on_start": false, + "use_liger_kernel": false, + "liger_kernel_config": null, + "eval_use_gather_object": false, + "average_tokens_across_devices": true, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": null, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "metric": [ + "temporal_grounding" + ], + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "Qwen/Qwen3-VL-4B-Instruct", + "model_type": "qwen3_vl", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": "flash_attn", + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_vl", + "system": null, + "max_length": 262144, + "truncation_strategy": "delete", + "max_pixels": null, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_free": false, + "padding_side": "right", + "loss_scale": "default", + "sequence_parallel_size": 4, + "response_prefix": null, + "template_backend": "swift", + "dataset": [ + "../data/timelens-100k/train/train.jsonl" + ], + "val_dataset": [], + "cached_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 1, + "load_from_cache_file": true, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "ckpt_dir": null, + "lora_modules": [], + "train_type": "full", + "adapters": [], + "external_plugins": [], + "model_kwargs": {}, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "custom_register_path": [], + "use_hf": true, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [ + "model.visual.merger", + "model.visual.deepstack_merger_list" + ], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": false, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": null, + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_lark_webhook_url": null, + "swanlab_lark_secret": null, + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "early_stop_interval": null, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3-VL-4B-Instruct", + "model_info": "ModelInfo(model_type='qwen3_vl', model_dir='/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling={'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default'}, is_moe_model=False, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_vl', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct', hf_model_id='Qwen/Qwen3-VL-2B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking', hf_model_id='Qwen/Qwen3-VL-2B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct', hf_model_id='Qwen/Qwen3-VL-4B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking', hf_model_id='Qwen/Qwen3-VL-4B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct', hf_model_id='Qwen/Qwen3-VL-8B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking', hf_model_id='Qwen/Qwen3-VL-8B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct', hf_model_id='Qwen/Qwen3-VL-32B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking', hf_model_id='Qwen/Qwen3-VL-32B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='qwen3_vl', get_function=, model_arch=MultiModelKeys(arch_name='qwen3_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model'], aligner=['model.visual.merger', 'model.visual.deepstack_merger_list'], vision_tower=['model.visual'], generator=[]), architectures=['Qwen3VLForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, is_reranker=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])", + "model_dir": "/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17", + "_val_dataset_exists": [], + "hub": "", + "evaluation_strategy": "epoch", + "training_args": "Seq2SeqTrainingArguments(output_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442', overwrite_output_dir=False, do_train=False, do_eval=False, do_predict=False, eval_strategy=, prediction_loss_only=False, per_device_train_batch_size=4, per_device_eval_batch_size=4, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=4, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=1e-05, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=1.0, max_steps=-1, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_ratio=0.03, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/runs', logging_strategy=, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=, save_steps=500, save_total_limit=2, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, jit_mode_eval=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=None, dataloader_num_workers=24, dataloader_prefetch_factor=2, past_index=-1, run_name='qwen3vl_4b_sft_timelens100k_1epoch', disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, deepspeed={'fp16': {'enabled': 'auto', 'loss_scale': 0, 'loss_scale_window': 1000, 'initial_scale_power': 16, 'hysteresis': 2, 'min_loss_scale': 1}, 'bf16': {'enabled': 'auto'}, 'zero_optimization': {'stage': 2, 'offload_optimizer': {'device': 'none', 'pin_memory': True}, 'allgather_partitions': True, 'allgather_bucket_size': 200000000.0, 'overlap_comm': False, 'reduce_scatter': True, 'reduce_bucket_size': 200000000.0, 'contiguous_gradients': True}, 'gradient_accumulation_steps': 'auto', 'gradient_clipping': 'auto', 'steps_per_print': 2000, 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'wall_clock_breakdown': False}, label_smoothing_factor=0.0, optim=, optim_args=None, adafactor=False, group_by_length=False, length_column_name='length', report_to=['wandb'], project='huggingface', trackio_space_id='trackio', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint=None, hub_model_id=None, hub_strategy=, hub_token=None, hub_private_repo=None, hub_always_push=False, hub_revision=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=18000000, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, include_tokens_per_second=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, liger_kernel_config=None, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=True, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, check_model=True, acc_strategy='token', train_dataloader_shuffle=True, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, metric=['temporal_grounding'], eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, use_flash_ckpt=False, sft_alpha=0, chord_sft_dataset=[], chord_sft_per_device_train_batch_size=None, chord_enable_phi_function=False, chord_mu_warmup_steps=None, chord_mu_decay_steps=None, chord_mu_peak=None, chord_mu_valley=None, train_type='full', local_repo_path=None, galore_config=None, padding_side='right', padding_free=False, task_type='causal_lm', problem_type=None)" +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/chat_template.jinja b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/config.json b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6cc71b039348c388053f379414d8d353760254f1 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/config.json @@ -0,0 +1,70 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "bfloat16", + "eos_token_id": 151645, + "hidden_size": 2560, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "bfloat16", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9728, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "pad_token_id": 151643, + "rms_norm_eps": 1e-06, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_type": "default" + }, + "rope_theta": 5000000, + "tie_word_embeddings": true, + "use_cache": false, + "vocab_size": 151936 + }, + "tie_word_embeddings": true, + "transformers_version": "4.57.0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 5, + 11, + 17 + ], + "depth": 24, + "dtype": "bfloat16", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "pad_token_id": 151643, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/generation_config.json b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..473d046ae2c09fde761a299ddc91ab5ba3cf67be --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "4.57.0" +} diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/latest b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/latest new file mode 100644 index 0000000000000000000000000000000000000000..3ada2ebdbf41759d59c713cc3f77afaa58a6b2aa --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/latest @@ -0,0 +1 @@ +global_step3019 \ No newline at end of file diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/model.safetensors.index.json b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..5185ad4158d213752a144bd6053b766e0262d5f2 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/model.safetensors.index.json @@ -0,0 +1,722 @@ +{ + "metadata": { + "total_parameters": 4437815808, + "total_size": 9653543936 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.language_model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.patch_embed.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.patch_embed.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.pos_embed.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/preprocessor_config.json b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..2ea84a437d448ff71b08df68fdd949d5cc4ebb64 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/preprocessor_config.json @@ -0,0 +1,21 @@ +{ + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "patch_size": 16, + "temporal_patch_size": 2, + "merge_size": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "processor_class": "Qwen3VLProcessor", + "image_processor_type": "Qwen2VLImageProcessorFast" +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_0.pth b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..4b0bea5e7a82c1886ccb79d111257405625494a6 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f8f92cf63e0989759370d24108b469c492c12202403f036015307ce49f12cedc +size 16389 diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_1.pth b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..2c94d69496797fa9c0f83f8a534c8ba692cb143a --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9ed40a0a4e9f365d2c6cc004d97e6705894eba46c8be4c160c1455bc3062dee1 +size 16389 diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_2.pth b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..e3d2f614c50841e8255840051884b16a6e22c38a --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d688b304d19c260b5cfa471535ed51d7e1d60b3a0d0159dfd1a04b87904a9f42 +size 16389 diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_3.pth b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_3.pth new file mode 100644 index 0000000000000000000000000000000000000000..84660ff39da893f3edf58f4a54971ca424b5ae70 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_3.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9967425ebcaee80d9b518fa0244d52f739b1b983d87cda71d5fede0c073e9d3b +size 16389 diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_4.pth b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_4.pth new file mode 100644 index 0000000000000000000000000000000000000000..fd63d0f91d990383a77a110da21707c2e4161995 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_4.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:469900fd39c667ffbd49c3c407c0ba317a1e9f5f9339a99b5d38423b7d0ce6d4 +size 16389 diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_5.pth b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_5.pth new file mode 100644 index 0000000000000000000000000000000000000000..a9f2a80a12ac3102dc70232f90338ece968e5537 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_5.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:124688471ff2a6e80f2fcefedbf741fb18d08dd539d5bd07a52e81be545142a5 +size 16389 diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_6.pth b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_6.pth new file mode 100644 index 0000000000000000000000000000000000000000..ffdeaa99e14137e82d32d685f2311f4ac57f4cd1 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_6.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1e69f1ced9f992a72c948698e5eb06088610788988cdb2fdbdd624e064319d60 +size 16389 diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_7.pth b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_7.pth new file mode 100644 index 0000000000000000000000000000000000000000..ad30a51583bfc3be86c6b6aa3bc2e815019b3e77 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/rng_state_7.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a376268a55d6ee10c371c06aa952334c4c6a1af9ea2d71b1951a57367a0c6722 +size 16389 diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/scheduler.pt b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..d8dc16a49cdcd12ce0c78aee169323978df7d6d5 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c52f04e4367e5fc71e3cb86228ecc3fe687958f8162c42728a816520eb513c7f +size 1465 diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/special_tokens_map.json b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/tokenizer_config.json b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/trainer_state.json b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..9c556fc6d63d354260fa9814f5f926cd1cb78bfc --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/trainer_state.json @@ -0,0 +1,7282 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 3019, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00033129037601457677, + "grad_norm": 37.5207633972168, + "learning_rate": 1.098901098901099e-07, + "loss": 3.310546875, + "step": 1, + "tIoU": 0.0, + "tRecall_50": 0.0, + "tRecall_75": 0.0, + "tRecall_90": 0.0, + "token_acc": 0.7585733882030178 + }, + { + "epoch": 0.001656451880072884, + "grad_norm": 33.392704010009766, + "learning_rate": 5.494505494505495e-07, + "loss": 3.28759765625, + "step": 5, + "tIoU": 0.0, + "tRecall_50": 0.0, + "tRecall_75": 0.0, + "tRecall_90": 0.0, + "token_acc": 0.7540308747855917 + }, + { + "epoch": 0.003312903760145768, + "grad_norm": 23.32330894470215, + "learning_rate": 1.098901098901099e-06, + "loss": 2.9546875, + "step": 10, + "tIoU": 0.0, + "tRecall_50": 0.0, + "tRecall_75": 0.0, + "tRecall_90": 0.0, + "token_acc": 0.779031370390754 + }, + { + "epoch": 0.0049693556402186515, + "grad_norm": 15.744415283203125, + "learning_rate": 1.6483516483516484e-06, + "loss": 1.830859375, + "step": 15, + "tIoU": 0.0, + "tRecall_50": 0.0, + "tRecall_75": 0.0, + "tRecall_90": 0.0, + "token_acc": 0.8237547892720306 + }, + { + "epoch": 0.006625807520291536, + "grad_norm": 13.483047485351562, + "learning_rate": 2.197802197802198e-06, + "loss": 0.6703125, + "step": 20, + "tIoU": 0.15047523975372315, + "tRecall_50": 0.15, + "tRecall_75": 0.09375, + "tRecall_90": 0.05, + "token_acc": 0.8428221193135386 + }, + { + "epoch": 0.008282259400364419, + "grad_norm": 1.4258636236190796, + "learning_rate": 2.7472527472527476e-06, + "loss": 0.21407470703125, + "step": 25, + "tIoU": 0.47949848175048826, + "tRecall_50": 0.5, + "tRecall_75": 0.2875, + "tRecall_90": 0.18125, + "token_acc": 0.9279303780255643 + }, + { + "epoch": 0.009938711280437303, + "grad_norm": 1.9763015508651733, + "learning_rate": 3.2967032967032968e-06, + "loss": 0.1758056640625, + "step": 30, + "tIoU": 0.5248600482940674, + "tRecall_50": 0.59375, + "tRecall_75": 0.35625, + "tRecall_90": 0.2125, + "token_acc": 0.9358974358974359 + }, + { + "epoch": 0.011595163160510187, + "grad_norm": 2.235086679458618, + "learning_rate": 3.846153846153847e-06, + "loss": 0.19344482421875, + "step": 35, + "tIoU": 0.4807941436767578, + "tRecall_50": 0.5125, + "tRecall_75": 0.33125, + "tRecall_90": 0.225, + "token_acc": 0.9296811120196239 + }, + { + "epoch": 0.013251615040583071, + "grad_norm": 1.306754231452942, + "learning_rate": 4.395604395604396e-06, + "loss": 0.1904052734375, + "step": 40, + "tIoU": 0.45671963691711426, + "tRecall_50": 0.475, + "tRecall_75": 0.275, + "tRecall_90": 0.15625, + "token_acc": 0.9292107404393816 + }, + { + "epoch": 0.014908066920655955, + "grad_norm": 1.1737031936645508, + "learning_rate": 4.945054945054946e-06, + "loss": 0.18565673828125, + "step": 45, + "tIoU": 0.4578300952911377, + "tRecall_50": 0.5, + "tRecall_75": 0.34375, + "tRecall_90": 0.24375, + "token_acc": 0.9360131255127153 + }, + { + "epoch": 0.016564518800728838, + "grad_norm": 1.6940656900405884, + "learning_rate": 5.494505494505495e-06, + "loss": 0.19178466796875, + "step": 50, + "tIoU": 0.44655814170837405, + "tRecall_50": 0.44375, + "tRecall_75": 0.31875, + "tRecall_90": 0.18125, + "token_acc": 0.9292543021032504 + }, + { + "epoch": 0.018220970680801724, + "grad_norm": 2.1192846298217773, + "learning_rate": 6.043956043956044e-06, + "loss": 0.18096923828125, + "step": 55, + "tIoU": 0.44701719284057617, + "tRecall_50": 0.5, + "tRecall_75": 0.31875, + "tRecall_90": 0.16875, + "token_acc": 0.9279824079164376 + }, + { + "epoch": 0.019877422560874606, + "grad_norm": 1.134567141532898, + "learning_rate": 6.5934065934065935e-06, + "loss": 0.210260009765625, + "step": 60, + "tIoU": 0.39896743297576903, + "tRecall_50": 0.4125, + "tRecall_75": 0.30625, + "tRecall_90": 0.19375, + "token_acc": 0.925845147219193 + }, + { + "epoch": 0.021533874440947492, + "grad_norm": 2.079129219055176, + "learning_rate": 7.1428571428571436e-06, + "loss": 0.19576416015625, + "step": 65, + "tIoU": 0.4517733097076416, + "tRecall_50": 0.51875, + "tRecall_75": 0.26875, + "tRecall_90": 0.14375, + "token_acc": 0.9273318872017353 + }, + { + "epoch": 0.023190326321020374, + "grad_norm": 2.3562278747558594, + "learning_rate": 7.692307692307694e-06, + "loss": 0.198193359375, + "step": 70, + "tIoU": 0.48453369140625, + "tRecall_50": 0.53125, + "tRecall_75": 0.35625, + "tRecall_90": 0.19375, + "token_acc": 0.9291166848418757 + }, + { + "epoch": 0.024846778201093257, + "grad_norm": 0.9555505514144897, + "learning_rate": 8.241758241758243e-06, + "loss": 0.1886962890625, + "step": 75, + "tIoU": 0.4445176601409912, + "tRecall_50": 0.4875, + "tRecall_75": 0.3375, + "tRecall_90": 0.1625, + "token_acc": 0.9272082878953108 + }, + { + "epoch": 0.026503230081166142, + "grad_norm": 1.0119189023971558, + "learning_rate": 8.791208791208792e-06, + "loss": 0.18936767578125, + "step": 80, + "tIoU": 0.4313098907470703, + "tRecall_50": 0.46875, + "tRecall_75": 0.25625, + "tRecall_90": 0.15, + "token_acc": 0.9278518921862238 + }, + { + "epoch": 0.028159681961239025, + "grad_norm": 1.2861628532409668, + "learning_rate": 9.340659340659341e-06, + "loss": 0.2029052734375, + "step": 85, + "tIoU": 0.42988147735595705, + "tRecall_50": 0.44375, + "tRecall_75": 0.3125, + "tRecall_90": 0.19375, + "token_acc": 0.9288451012588944 + }, + { + "epoch": 0.02981613384131191, + "grad_norm": 0.9592366218566895, + "learning_rate": 9.890109890109892e-06, + "loss": 0.18587646484375, + "step": 90, + "tIoU": 0.4889820098876953, + "tRecall_50": 0.525, + "tRecall_75": 0.36875, + "tRecall_90": 0.1875, + "token_acc": 0.9345156889495225 + }, + { + "epoch": 0.031472585721384796, + "grad_norm": 1.294869303703308, + "learning_rate": 9.99995395134759e-06, + "loss": 0.18406982421875, + "step": 95, + "tIoU": 0.44582033157348633, + "tRecall_50": 0.49375, + "tRecall_75": 0.30625, + "tRecall_90": 0.14375, + "token_acc": 0.9277762693456422 + }, + { + "epoch": 0.033129037601457675, + "grad_norm": 1.088860034942627, + "learning_rate": 9.999766880150859e-06, + "loss": 0.209716796875, + "step": 100, + "tIoU": 0.4009531497955322, + "tRecall_50": 0.425, + "tRecall_75": 0.21875, + "tRecall_90": 0.1375, + "token_acc": 0.9226368836829202 + }, + { + "epoch": 0.03478548948153056, + "grad_norm": 0.7980255484580994, + "learning_rate": 9.99943591374887e-06, + "loss": 0.210595703125, + "step": 105, + "tIoU": 0.42629294395446776, + "tRecall_50": 0.4625, + "tRecall_75": 0.275, + "tRecall_90": 0.10625, + "token_acc": 0.9228687415426252 + }, + { + "epoch": 0.03644194136160345, + "grad_norm": 1.2525278329849243, + "learning_rate": 9.998961061666965e-06, + "loss": 0.20194091796875, + "step": 110, + "tIoU": 0.4068713665008545, + "tRecall_50": 0.4375, + "tRecall_75": 0.2375, + "tRecall_90": 0.14375, + "token_acc": 0.9252514270182115 + }, + { + "epoch": 0.038098393241676326, + "grad_norm": 1.621291995048523, + "learning_rate": 9.998342337571566e-06, + "loss": 0.2082275390625, + "step": 115, + "tIoU": 0.4473285675048828, + "tRecall_50": 0.50625, + "tRecall_75": 0.2625, + "tRecall_90": 0.14375, + "token_acc": 0.9251904243743199 + }, + { + "epoch": 0.03975484512174921, + "grad_norm": 0.8591983914375305, + "learning_rate": 9.997579759269792e-06, + "loss": 0.2082763671875, + "step": 120, + "tIoU": 0.419110631942749, + "tRecall_50": 0.44375, + "tRecall_75": 0.29375, + "tRecall_90": 0.15625, + "token_acc": 0.92253136933988 + }, + { + "epoch": 0.0414112970018221, + "grad_norm": 4.022842884063721, + "learning_rate": 9.996673348708936e-06, + "loss": 0.203955078125, + "step": 125, + "tIoU": 0.41852197647094724, + "tRecall_50": 0.4375, + "tRecall_75": 0.26875, + "tRecall_90": 0.14375, + "token_acc": 0.9234527687296417 + }, + { + "epoch": 0.043067748881894984, + "grad_norm": 1.0151914358139038, + "learning_rate": 9.995623131975842e-06, + "loss": 0.2289794921875, + "step": 130, + "tIoU": 0.3222654342651367, + "tRecall_50": 0.35, + "tRecall_75": 0.21875, + "tRecall_90": 0.125, + "token_acc": 0.9190579317812669 + }, + { + "epoch": 0.04472420076196786, + "grad_norm": 0.762208104133606, + "learning_rate": 9.99442913929615e-06, + "loss": 0.2308349609375, + "step": 135, + "tIoU": 0.348409628868103, + "tRecall_50": 0.375, + "tRecall_75": 0.2, + "tRecall_90": 0.10625, + "token_acc": 0.9160076356694846 + }, + { + "epoch": 0.04638065264204075, + "grad_norm": 1.0656577348709106, + "learning_rate": 9.993091405033427e-06, + "loss": 0.2153564453125, + "step": 140, + "tIoU": 0.34590022563934325, + "tRecall_50": 0.34375, + "tRecall_75": 0.21875, + "tRecall_90": 0.15625, + "token_acc": 0.9209518599562363 + }, + { + "epoch": 0.048037104522113634, + "grad_norm": 1.2896987199783325, + "learning_rate": 9.991609967688177e-06, + "loss": 0.1908203125, + "step": 145, + "tIoU": 0.468013858795166, + "tRecall_50": 0.5125, + "tRecall_75": 0.31875, + "tRecall_90": 0.18125, + "token_acc": 0.9322774440196614 + }, + { + "epoch": 0.04969355640218651, + "grad_norm": 0.8637948036193848, + "learning_rate": 9.989984869896734e-06, + "loss": 0.2052490234375, + "step": 150, + "tIoU": 0.43006191253662107, + "tRecall_50": 0.43125, + "tRecall_75": 0.2875, + "tRecall_90": 0.14375, + "token_acc": 0.9274017467248908 + }, + { + "epoch": 0.0513500082822594, + "grad_norm": 1.6045029163360596, + "learning_rate": 9.988216158430033e-06, + "loss": 0.18021240234375, + "step": 155, + "tIoU": 0.5069736480712891, + "tRecall_50": 0.55625, + "tRecall_75": 0.4, + "tRecall_90": 0.21875, + "token_acc": 0.9343860604410563 + }, + { + "epoch": 0.053006460162332285, + "grad_norm": 1.5454106330871582, + "learning_rate": 9.986303884192269e-06, + "loss": 0.19678955078125, + "step": 160, + "tIoU": 0.44966754913330076, + "tRecall_50": 0.5, + "tRecall_75": 0.3125, + "tRecall_90": 0.16875, + "token_acc": 0.926550598476605 + }, + { + "epoch": 0.05466291204240517, + "grad_norm": 0.8348360657691956, + "learning_rate": 9.984248102219428e-06, + "loss": 0.224371337890625, + "step": 165, + "tIoU": 0.4493142604827881, + "tRecall_50": 0.475, + "tRecall_75": 0.30625, + "tRecall_90": 0.1875, + "token_acc": 0.9201303993480032 + }, + { + "epoch": 0.05631936392247805, + "grad_norm": 1.0888898372650146, + "learning_rate": 9.9820488716777e-06, + "loss": 0.199365234375, + "step": 170, + "tIoU": 0.434508752822876, + "tRecall_50": 0.45625, + "tRecall_75": 0.2625, + "tRecall_90": 0.13125, + "token_acc": 0.9254880395930712 + }, + { + "epoch": 0.057975815802550935, + "grad_norm": 1.4940712451934814, + "learning_rate": 9.97970625586178e-06, + "loss": 0.1918701171875, + "step": 175, + "tIoU": 0.4550649166107178, + "tRecall_50": 0.49375, + "tRecall_75": 0.3375, + "tRecall_90": 0.15625, + "token_acc": 0.9296427597491137 + }, + { + "epoch": 0.05963226768262382, + "grad_norm": 0.890186071395874, + "learning_rate": 9.977220322193056e-06, + "loss": 0.2085205078125, + "step": 180, + "tIoU": 0.3760667324066162, + "tRecall_50": 0.40625, + "tRecall_75": 0.24375, + "tRecall_90": 0.11875, + "token_acc": 0.9240851993446204 + }, + { + "epoch": 0.06128871956269671, + "grad_norm": 0.7507341504096985, + "learning_rate": 9.974591142217643e-06, + "loss": 0.19017333984375, + "step": 185, + "tIoU": 0.46780953407287595, + "tRecall_50": 0.525, + "tRecall_75": 0.3375, + "tRecall_90": 0.15, + "token_acc": 0.9292349726775956 + }, + { + "epoch": 0.06294517144276959, + "grad_norm": 1.2498581409454346, + "learning_rate": 9.971818791604352e-06, + "loss": 0.23975830078125, + "step": 190, + "tIoU": 0.37864179611206056, + "tRecall_50": 0.38125, + "tRecall_75": 0.24375, + "tRecall_90": 0.125, + "token_acc": 0.9137884144683166 + }, + { + "epoch": 0.06460162332284247, + "grad_norm": 1.1122004985809326, + "learning_rate": 9.968903350142494e-06, + "loss": 0.192022705078125, + "step": 195, + "tIoU": 0.4604650020599365, + "tRecall_50": 0.46875, + "tRecall_75": 0.3375, + "tRecall_90": 0.2, + "token_acc": 0.9331331877729258 + }, + { + "epoch": 0.06625807520291535, + "grad_norm": 0.8076160550117493, + "learning_rate": 9.965844901739587e-06, + "loss": 0.2272705078125, + "step": 200, + "tIoU": 0.32807245254516604, + "tRecall_50": 0.35, + "tRecall_75": 0.2, + "tRecall_90": 0.14375, + "token_acc": 0.9176663031624863 + }, + { + "epoch": 0.06791452708298824, + "grad_norm": 1.5814825296401978, + "learning_rate": 9.962643534418954e-06, + "loss": 0.2008056640625, + "step": 205, + "tIoU": 0.4125843524932861, + "tRecall_50": 0.475, + "tRecall_75": 0.2375, + "tRecall_90": 0.15, + "token_acc": 0.9271089271089271 + }, + { + "epoch": 0.06957097896306112, + "grad_norm": 0.9881187677383423, + "learning_rate": 9.959299340317167e-06, + "loss": 0.19127197265625, + "step": 210, + "tIoU": 0.4517991065979004, + "tRecall_50": 0.5, + "tRecall_75": 0.28125, + "tRecall_90": 0.13125, + "token_acc": 0.9266104919815167 + }, + { + "epoch": 0.071227430843134, + "grad_norm": 0.862718403339386, + "learning_rate": 9.955812415681415e-06, + "loss": 0.20115966796875, + "step": 215, + "tIoU": 0.43936910629272463, + "tRecall_50": 0.44375, + "tRecall_75": 0.26875, + "tRecall_90": 0.15625, + "token_acc": 0.925531914893617 + }, + { + "epoch": 0.0728838827232069, + "grad_norm": 0.8651227951049805, + "learning_rate": 9.952182860866723e-06, + "loss": 0.206103515625, + "step": 220, + "tIoU": 0.4165684700012207, + "tRecall_50": 0.4125, + "tRecall_75": 0.23125, + "tRecall_90": 0.1375, + "token_acc": 0.9257033597377765 + }, + { + "epoch": 0.07454033460327977, + "grad_norm": 0.7443090081214905, + "learning_rate": 9.94841078033307e-06, + "loss": 0.20802001953125, + "step": 225, + "tIoU": 0.41522841453552245, + "tRecall_50": 0.425, + "tRecall_75": 0.30625, + "tRecall_90": 0.1625, + "token_acc": 0.924877517691889 + }, + { + "epoch": 0.07619678648335265, + "grad_norm": 0.6482592821121216, + "learning_rate": 9.944496282642377e-06, + "loss": 0.19639892578125, + "step": 230, + "tIoU": 0.4244387626647949, + "tRecall_50": 0.45, + "tRecall_75": 0.3, + "tRecall_90": 0.1625, + "token_acc": 0.9295890410958904 + }, + { + "epoch": 0.07785323836342554, + "grad_norm": 0.8621935248374939, + "learning_rate": 9.940439480455386e-06, + "loss": 0.1994873046875, + "step": 235, + "tIoU": 0.43390598297119143, + "tRecall_50": 0.4625, + "tRecall_75": 0.28125, + "tRecall_90": 0.14375, + "token_acc": 0.9281842818428184 + }, + { + "epoch": 0.07950969024349842, + "grad_norm": 0.9681967496871948, + "learning_rate": 9.936240490528419e-06, + "loss": 0.2104248046875, + "step": 240, + "tIoU": 0.3532999277114868, + "tRecall_50": 0.35, + "tRecall_75": 0.2, + "tRecall_90": 0.1125, + "token_acc": 0.9234345091605141 + }, + { + "epoch": 0.08116614212357132, + "grad_norm": 2.0264437198638916, + "learning_rate": 9.931899433710011e-06, + "loss": 0.2027099609375, + "step": 245, + "tIoU": 0.40817694664001464, + "tRecall_50": 0.43125, + "tRecall_75": 0.275, + "tRecall_90": 0.2, + "token_acc": 0.924863387978142 + }, + { + "epoch": 0.0828225940036442, + "grad_norm": 1.126722812652588, + "learning_rate": 9.927416434937442e-06, + "loss": 0.2103515625, + "step": 250, + "tIoU": 0.4178160190582275, + "tRecall_50": 0.4625, + "tRecall_75": 0.2875, + "tRecall_90": 0.1625, + "token_acc": 0.9214285714285714 + }, + { + "epoch": 0.08447904588371707, + "grad_norm": 0.7986385226249695, + "learning_rate": 9.922791623233134e-06, + "loss": 0.20733642578125, + "step": 255, + "tIoU": 0.45475945472717283, + "tRecall_50": 0.49375, + "tRecall_75": 0.30625, + "tRecall_90": 0.20625, + "token_acc": 0.9269493844049248 + }, + { + "epoch": 0.08613549776378997, + "grad_norm": 0.7802565693855286, + "learning_rate": 9.918025131700935e-06, + "loss": 0.2189697265625, + "step": 260, + "tIoU": 0.390622615814209, + "tRecall_50": 0.43125, + "tRecall_75": 0.225, + "tRecall_90": 0.125, + "token_acc": 0.9224511930585684 + }, + { + "epoch": 0.08779194964386285, + "grad_norm": 0.6944566369056702, + "learning_rate": 9.9131170975223e-06, + "loss": 0.21519775390625, + "step": 265, + "tIoU": 0.37258186340332033, + "tRecall_50": 0.40625, + "tRecall_75": 0.23125, + "tRecall_90": 0.1125, + "token_acc": 0.9179165530406327 + }, + { + "epoch": 0.08944840152393572, + "grad_norm": 0.7498219609260559, + "learning_rate": 9.90806766195233e-06, + "loss": 0.23343505859375, + "step": 270, + "tIoU": 0.3832035779953003, + "tRecall_50": 0.39375, + "tRecall_75": 0.24375, + "tRecall_90": 0.1375, + "token_acc": 0.913684785577711 + }, + { + "epoch": 0.09110485340400862, + "grad_norm": 0.996425211429596, + "learning_rate": 9.902876970315715e-06, + "loss": 0.20540771484375, + "step": 275, + "tIoU": 0.44122681617736814, + "tRecall_50": 0.5, + "tRecall_75": 0.25, + "tRecall_90": 0.1375, + "token_acc": 0.9231607629427793 + }, + { + "epoch": 0.0927613052840815, + "grad_norm": 0.9214315414428711, + "learning_rate": 9.897545172002547e-06, + "loss": 0.2068359375, + "step": 280, + "tIoU": 0.42021684646606444, + "tRecall_50": 0.4375, + "tRecall_75": 0.29375, + "tRecall_90": 0.1625, + "token_acc": 0.9262295081967213 + }, + { + "epoch": 0.09441775716415438, + "grad_norm": 3.090954065322876, + "learning_rate": 9.892072420464026e-06, + "loss": 0.197119140625, + "step": 285, + "tIoU": 0.43810601234436036, + "tRecall_50": 0.48125, + "tRecall_75": 0.2875, + "tRecall_90": 0.16875, + "token_acc": 0.9290605313612709 + }, + { + "epoch": 0.09607420904422727, + "grad_norm": 1.0458402633666992, + "learning_rate": 9.886458873208032e-06, + "loss": 0.20242919921875, + "step": 290, + "tIoU": 0.44785561561584475, + "tRecall_50": 0.5, + "tRecall_75": 0.30625, + "tRecall_90": 0.1875, + "token_acc": 0.9299154622307063 + }, + { + "epoch": 0.09773066092430015, + "grad_norm": 0.8456872701644897, + "learning_rate": 9.880704691794608e-06, + "loss": 0.2062255859375, + "step": 295, + "tIoU": 0.42615447044372556, + "tRecall_50": 0.45, + "tRecall_75": 0.3, + "tRecall_90": 0.16875, + "token_acc": 0.9247135842880524 + }, + { + "epoch": 0.09938711280437303, + "grad_norm": 0.9653769135475159, + "learning_rate": 9.874810041831298e-06, + "loss": 0.20479736328125, + "step": 300, + "tIoU": 0.39232654571533204, + "tRecall_50": 0.4125, + "tRecall_75": 0.28125, + "tRecall_90": 0.14375, + "token_acc": 0.9242053789731052 + }, + { + "epoch": 0.10104356468444592, + "grad_norm": 0.8343780636787415, + "learning_rate": 9.868775092968384e-06, + "loss": 0.1989013671875, + "step": 305, + "tIoU": 0.41749234199523927, + "tRecall_50": 0.4375, + "tRecall_75": 0.3125, + "tRecall_90": 0.15, + "token_acc": 0.9305402425578831 + }, + { + "epoch": 0.1027000165645188, + "grad_norm": 0.815432608127594, + "learning_rate": 9.862600018894004e-06, + "loss": 0.19197998046875, + "step": 310, + "tIoU": 0.4726102828979492, + "tRecall_50": 0.55, + "tRecall_75": 0.3125, + "tRecall_90": 0.15, + "token_acc": 0.9284158954817637 + }, + { + "epoch": 0.10435646844459169, + "grad_norm": 0.6193230152130127, + "learning_rate": 9.856284997329158e-06, + "loss": 0.20172119140625, + "step": 315, + "tIoU": 0.4494438171386719, + "tRecall_50": 0.49375, + "tRecall_75": 0.3375, + "tRecall_90": 0.1875, + "token_acc": 0.9278350515463918 + }, + { + "epoch": 0.10601292032466457, + "grad_norm": 0.7088842391967773, + "learning_rate": 9.849830210022582e-06, + "loss": 0.20203857421875, + "step": 320, + "tIoU": 0.43042378425598143, + "tRecall_50": 0.45625, + "tRecall_75": 0.29375, + "tRecall_90": 0.15625, + "token_acc": 0.9257748776508973 + }, + { + "epoch": 0.10766937220473745, + "grad_norm": 0.6478131413459778, + "learning_rate": 9.843235842745527e-06, + "loss": 0.196875, + "step": 325, + "tIoU": 0.45046267509460447, + "tRecall_50": 0.5, + "tRecall_75": 0.3375, + "tRecall_90": 0.15625, + "token_acc": 0.9284739284739285 + }, + { + "epoch": 0.10932582408481034, + "grad_norm": 0.5544605851173401, + "learning_rate": 9.836502085286408e-06, + "loss": 0.20408935546875, + "step": 330, + "tIoU": 0.47177977561950685, + "tRecall_50": 0.5, + "tRecall_75": 0.33125, + "tRecall_90": 0.19375, + "token_acc": 0.9275599128540305 + }, + { + "epoch": 0.11098227596488322, + "grad_norm": 0.7553025484085083, + "learning_rate": 9.829629131445342e-06, + "loss": 0.202001953125, + "step": 335, + "tIoU": 0.424683141708374, + "tRecall_50": 0.45, + "tRecall_75": 0.25625, + "tRecall_90": 0.1375, + "token_acc": 0.9231816943612094 + }, + { + "epoch": 0.1126387278449561, + "grad_norm": 24.649581909179688, + "learning_rate": 9.822617179028575e-06, + "loss": 0.3250732421875, + "step": 340, + "tIoU": 0.33823087215423586, + "tRecall_50": 0.3625, + "tRecall_75": 0.23125, + "tRecall_90": 0.16875, + "token_acc": 0.9176630434782609 + }, + { + "epoch": 0.11429517972502899, + "grad_norm": 1.2221829891204834, + "learning_rate": 9.815466429842781e-06, + "loss": 0.19600830078125, + "step": 345, + "tIoU": 0.4647655487060547, + "tRecall_50": 0.51875, + "tRecall_75": 0.34375, + "tRecall_90": 0.18125, + "token_acc": 0.9318860244233379 + }, + { + "epoch": 0.11595163160510187, + "grad_norm": 0.5191197395324707, + "learning_rate": 9.808177089689258e-06, + "loss": 0.19267578125, + "step": 350, + "tIoU": 0.444231653213501, + "tRecall_50": 0.49375, + "tRecall_75": 0.28125, + "tRecall_90": 0.1375, + "token_acc": 0.9275834011391375 + }, + { + "epoch": 0.11760808348517475, + "grad_norm": 0.7791886925697327, + "learning_rate": 9.80074936835801e-06, + "loss": 0.19835205078125, + "step": 355, + "tIoU": 0.43950557708740234, + "tRecall_50": 0.425, + "tRecall_75": 0.2875, + "tRecall_90": 0.15, + "token_acc": 0.9243789243789243 + }, + { + "epoch": 0.11926453536524764, + "grad_norm": 0.6446515917778015, + "learning_rate": 9.793183479621698e-06, + "loss": 0.2083251953125, + "step": 360, + "tIoU": 0.4317431926727295, + "tRecall_50": 0.45625, + "tRecall_75": 0.25625, + "tRecall_90": 0.15625, + "token_acc": 0.9247691472026073 + }, + { + "epoch": 0.12092098724532052, + "grad_norm": 0.6463059186935425, + "learning_rate": 9.7854796412295e-06, + "loss": 0.19989013671875, + "step": 365, + "tIoU": 0.4000588893890381, + "tRecall_50": 0.43125, + "tRecall_75": 0.21875, + "tRecall_90": 0.125, + "token_acc": 0.9270691068014204 + }, + { + "epoch": 0.12257743912539341, + "grad_norm": 0.8697914481163025, + "learning_rate": 9.777638074900831e-06, + "loss": 0.20377197265625, + "step": 370, + "tIoU": 0.38347530364990234, + "tRecall_50": 0.4, + "tRecall_75": 0.26875, + "tRecall_90": 0.1375, + "token_acc": 0.9237913138486752 + }, + { + "epoch": 0.1242338910054663, + "grad_norm": 0.8284282684326172, + "learning_rate": 9.769659006318976e-06, + "loss": 0.19049072265625, + "step": 375, + "tIoU": 0.41577610969543455, + "tRecall_50": 0.43125, + "tRecall_75": 0.30625, + "tRecall_90": 0.2125, + "token_acc": 0.933115823817292 + }, + { + "epoch": 0.12589034288553919, + "grad_norm": 0.544400155544281, + "learning_rate": 9.761542665124581e-06, + "loss": 0.19884033203125, + "step": 380, + "tIoU": 0.4304636001586914, + "tRecall_50": 0.48125, + "tRecall_75": 0.275, + "tRecall_90": 0.1625, + "token_acc": 0.9262697979246314 + }, + { + "epoch": 0.12754679476561206, + "grad_norm": 0.7749924659729004, + "learning_rate": 9.753289284909058e-06, + "loss": 0.196875, + "step": 385, + "tIoU": 0.45472006797790526, + "tRecall_50": 0.4625, + "tRecall_75": 0.3, + "tRecall_90": 0.18125, + "token_acc": 0.9281437125748503 + }, + { + "epoch": 0.12920324664568494, + "grad_norm": 0.8247610330581665, + "learning_rate": 9.744899103207844e-06, + "loss": 0.189599609375, + "step": 390, + "tIoU": 0.46105518341064455, + "tRecall_50": 0.45, + "tRecall_75": 0.33125, + "tRecall_90": 0.16875, + "token_acc": 0.9312328767123288 + }, + { + "epoch": 0.13085969852575782, + "grad_norm": 1.2671105861663818, + "learning_rate": 9.736372361493584e-06, + "loss": 0.2018798828125, + "step": 395, + "tIoU": 0.41598782539367674, + "tRecall_50": 0.4625, + "tRecall_75": 0.2625, + "tRecall_90": 0.125, + "token_acc": 0.9266104919815167 + }, + { + "epoch": 0.1325161504058307, + "grad_norm": 0.7221733331680298, + "learning_rate": 9.72770930516917e-06, + "loss": 0.20133056640625, + "step": 400, + "tIoU": 0.43442955017089846, + "tRecall_50": 0.43125, + "tRecall_75": 0.275, + "tRecall_90": 0.1625, + "token_acc": 0.9219219219219219 + }, + { + "epoch": 0.1341726022859036, + "grad_norm": 0.8075146079063416, + "learning_rate": 9.718910183560681e-06, + "loss": 0.19154052734375, + "step": 405, + "tIoU": 0.4739804744720459, + "tRecall_50": 0.53125, + "tRecall_75": 0.31875, + "tRecall_90": 0.14375, + "token_acc": 0.9313725490196079 + }, + { + "epoch": 0.1358290541659765, + "grad_norm": 0.6628204584121704, + "learning_rate": 9.709975249910203e-06, + "loss": 0.18717041015625, + "step": 410, + "tIoU": 0.48113312721252444, + "tRecall_50": 0.51875, + "tRecall_75": 0.33125, + "tRecall_90": 0.20625, + "token_acc": 0.9319486198414867 + }, + { + "epoch": 0.13748550604604937, + "grad_norm": 0.7093484401702881, + "learning_rate": 9.70090476136855e-06, + "loss": 0.2063232421875, + "step": 415, + "tIoU": 0.41611781120300295, + "tRecall_50": 0.4625, + "tRecall_75": 0.275, + "tRecall_90": 0.125, + "token_acc": 0.9220920730046309 + }, + { + "epoch": 0.13914195792612225, + "grad_norm": 0.5163065791130066, + "learning_rate": 9.691698978987855e-06, + "loss": 0.18984375, + "step": 420, + "tIoU": 0.45911922454833987, + "tRecall_50": 0.4875, + "tRecall_75": 0.30625, + "tRecall_90": 0.19375, + "token_acc": 0.931214031241436 + }, + { + "epoch": 0.14079840980619512, + "grad_norm": 0.6479618549346924, + "learning_rate": 9.68235816771406e-06, + "loss": 0.1943115234375, + "step": 425, + "tIoU": 0.41985330581665037, + "tRecall_50": 0.45625, + "tRecall_75": 0.2375, + "tRecall_90": 0.1125, + "token_acc": 0.9246873300706906 + }, + { + "epoch": 0.142454861686268, + "grad_norm": 1.3915655612945557, + "learning_rate": 9.672882596379289e-06, + "loss": 0.19625244140625, + "step": 430, + "tIoU": 0.4425868034362793, + "tRecall_50": 0.46875, + "tRecall_75": 0.24375, + "tRecall_90": 0.125, + "token_acc": 0.9271739130434783 + }, + { + "epoch": 0.1441113135663409, + "grad_norm": 0.874627947807312, + "learning_rate": 9.663272537694112e-06, + "loss": 0.196826171875, + "step": 435, + "tIoU": 0.4497288703918457, + "tRecall_50": 0.45, + "tRecall_75": 0.31875, + "tRecall_90": 0.1625, + "token_acc": 0.9272677744483792 + }, + { + "epoch": 0.1457677654464138, + "grad_norm": 0.6494178771972656, + "learning_rate": 9.653528268239703e-06, + "loss": 0.21339111328125, + "step": 440, + "tIoU": 0.3722111940383911, + "tRecall_50": 0.36875, + "tRecall_75": 0.225, + "tRecall_90": 0.15, + "token_acc": 0.9220708446866485 + }, + { + "epoch": 0.14742421732648667, + "grad_norm": 0.6357215642929077, + "learning_rate": 9.643650068459863e-06, + "loss": 0.18795166015625, + "step": 445, + "tIoU": 0.4609790325164795, + "tRecall_50": 0.50625, + "tRecall_75": 0.30625, + "tRecall_90": 0.16875, + "token_acc": 0.9316888045540797 + }, + { + "epoch": 0.14908066920655955, + "grad_norm": 0.9134014844894409, + "learning_rate": 9.633638222652967e-06, + "loss": 0.19854736328125, + "step": 450, + "tIoU": 0.4400688648223877, + "tRecall_50": 0.5, + "tRecall_75": 0.28125, + "tRecall_90": 0.15, + "token_acc": 0.9271884374147805 + }, + { + "epoch": 0.15073712108663243, + "grad_norm": 1.318121314048767, + "learning_rate": 9.62349301896377e-06, + "loss": 0.18902587890625, + "step": 455, + "tIoU": 0.4806161880493164, + "tRecall_50": 0.55, + "tRecall_75": 0.31875, + "tRecall_90": 0.1875, + "token_acc": 0.9309967141292442 + }, + { + "epoch": 0.1523935729667053, + "grad_norm": 0.8935189247131348, + "learning_rate": 9.613214749375117e-06, + "loss": 0.200537109375, + "step": 460, + "tIoU": 0.43247547149658205, + "tRecall_50": 0.45625, + "tRecall_75": 0.24375, + "tRecall_90": 0.1375, + "token_acc": 0.9250476969201418 + }, + { + "epoch": 0.1540500248467782, + "grad_norm": 0.8290932178497314, + "learning_rate": 9.602803709699544e-06, + "loss": 0.21024169921875, + "step": 465, + "tIoU": 0.4437929630279541, + "tRecall_50": 0.4625, + "tRecall_75": 0.275, + "tRecall_90": 0.1875, + "token_acc": 0.9262295081967213 + }, + { + "epoch": 0.1557064767268511, + "grad_norm": 0.7094597816467285, + "learning_rate": 9.592260199570761e-06, + "loss": 0.19241943359375, + "step": 470, + "tIoU": 0.4886190891265869, + "tRecall_50": 0.53125, + "tRecall_75": 0.30625, + "tRecall_90": 0.19375, + "token_acc": 0.9312414733969986 + }, + { + "epoch": 0.15736292860692397, + "grad_norm": 0.5024470090866089, + "learning_rate": 9.581584522435025e-06, + "loss": 0.19324951171875, + "step": 475, + "tIoU": 0.4388664722442627, + "tRecall_50": 0.4875, + "tRecall_75": 0.30625, + "tRecall_90": 0.1875, + "token_acc": 0.9317563893420338 + }, + { + "epoch": 0.15901938048699685, + "grad_norm": 0.7557206153869629, + "learning_rate": 9.570776985542412e-06, + "loss": 0.2156982421875, + "step": 480, + "tIoU": 0.4001711368560791, + "tRecall_50": 0.43125, + "tRecall_75": 0.25625, + "tRecall_90": 0.1375, + "token_acc": 0.9247751430907605 + }, + { + "epoch": 0.16067583236706973, + "grad_norm": 0.8227375149726868, + "learning_rate": 9.559837899937969e-06, + "loss": 0.189794921875, + "step": 485, + "tIoU": 0.47867345809936523, + "tRecall_50": 0.50625, + "tRecall_75": 0.3375, + "tRecall_90": 0.16875, + "token_acc": 0.9310250817884406 + }, + { + "epoch": 0.16233228424714263, + "grad_norm": 0.920362114906311, + "learning_rate": 9.548767580452774e-06, + "loss": 0.19400634765625, + "step": 490, + "tIoU": 0.47590069770812987, + "tRecall_50": 0.5375, + "tRecall_75": 0.31875, + "tRecall_90": 0.18125, + "token_acc": 0.9307441607821836 + }, + { + "epoch": 0.1639887361272155, + "grad_norm": 0.8449790477752686, + "learning_rate": 9.537566345694858e-06, + "loss": 0.1809326171875, + "step": 495, + "tIoU": 0.478501033782959, + "tRecall_50": 0.5, + "tRecall_75": 0.34375, + "tRecall_90": 0.20625, + "token_acc": 0.9327342047930284 + }, + { + "epoch": 0.1656451880072884, + "grad_norm": 0.7748955488204956, + "learning_rate": 9.526234518040045e-06, + "loss": 0.21148681640625, + "step": 500, + "tIoU": 0.43302106857299805, + "tRecall_50": 0.45, + "tRecall_75": 0.28125, + "tRecall_90": 0.175, + "token_acc": 0.9221980413492927 + }, + { + "epoch": 0.16730163988736127, + "grad_norm": 0.6476784944534302, + "learning_rate": 9.514772423622675e-06, + "loss": 0.19722900390625, + "step": 505, + "tIoU": 0.4332547664642334, + "tRecall_50": 0.46875, + "tRecall_75": 0.28125, + "tRecall_90": 0.1375, + "token_acc": 0.928298800436205 + }, + { + "epoch": 0.16895809176743415, + "grad_norm": 0.7341322898864746, + "learning_rate": 9.503180392326216e-06, + "loss": 0.19720458984375, + "step": 510, + "tIoU": 0.47381792068481443, + "tRecall_50": 0.4875, + "tRecall_75": 0.31875, + "tRecall_90": 0.19375, + "token_acc": 0.9332971800433839 + }, + { + "epoch": 0.17061454364750703, + "grad_norm": 0.9431743621826172, + "learning_rate": 9.491458757773766e-06, + "loss": 0.192437744140625, + "step": 515, + "tIoU": 0.4060690402984619, + "tRecall_50": 0.43125, + "tRecall_75": 0.2875, + "tRecall_90": 0.15625, + "token_acc": 0.92718313714755 + }, + { + "epoch": 0.17227099552757993, + "grad_norm": 0.9373649954795837, + "learning_rate": 9.479607857318454e-06, + "loss": 0.19864501953125, + "step": 520, + "tIoU": 0.4573049068450928, + "tRecall_50": 0.4875, + "tRecall_75": 0.325, + "tRecall_90": 0.16875, + "token_acc": 0.9318305268875611 + }, + { + "epoch": 0.1739274474076528, + "grad_norm": 0.6910871863365173, + "learning_rate": 9.467628032033736e-06, + "loss": 0.19825439453125, + "step": 525, + "tIoU": 0.4477407455444336, + "tRecall_50": 0.45625, + "tRecall_75": 0.33125, + "tRecall_90": 0.16875, + "token_acc": 0.9314505072662462 + }, + { + "epoch": 0.1755838992877257, + "grad_norm": 0.495835542678833, + "learning_rate": 9.455519626703569e-06, + "loss": 0.199853515625, + "step": 530, + "tIoU": 0.4348471164703369, + "tRecall_50": 0.45625, + "tRecall_75": 0.3, + "tRecall_90": 0.19375, + "token_acc": 0.9249796140255504 + }, + { + "epoch": 0.17724035116779857, + "grad_norm": 0.5815764665603638, + "learning_rate": 9.443282989812494e-06, + "loss": 0.19815673828125, + "step": 535, + "tIoU": 0.41817460060119627, + "tRecall_50": 0.44375, + "tRecall_75": 0.2625, + "tRecall_90": 0.15, + "token_acc": 0.9251700680272109 + }, + { + "epoch": 0.17889680304787145, + "grad_norm": 0.834153413772583, + "learning_rate": 9.430918473535608e-06, + "loss": 0.18758544921875, + "step": 540, + "tIoU": 0.41258726119995115, + "tRecall_50": 0.45625, + "tRecall_75": 0.28125, + "tRecall_90": 0.15625, + "token_acc": 0.925734494015234 + }, + { + "epoch": 0.18055325492794436, + "grad_norm": 0.49157267808914185, + "learning_rate": 9.418426433728422e-06, + "loss": 0.1734130859375, + "step": 545, + "tIoU": 0.49922776222229004, + "tRecall_50": 0.55, + "tRecall_75": 0.3875, + "tRecall_90": 0.2375, + "token_acc": 0.9362452523060227 + }, + { + "epoch": 0.18220970680801724, + "grad_norm": 0.8604575991630554, + "learning_rate": 9.405807229916627e-06, + "loss": 0.20540771484375, + "step": 550, + "tIoU": 0.40863842964172364, + "tRecall_50": 0.4375, + "tRecall_75": 0.28125, + "tRecall_90": 0.1875, + "token_acc": 0.9271232876712329 + }, + { + "epoch": 0.18386615868809011, + "grad_norm": 0.8063392043113708, + "learning_rate": 9.393061225285743e-06, + "loss": 0.18697509765625, + "step": 555, + "tIoU": 0.4734005928039551, + "tRecall_50": 0.53125, + "tRecall_75": 0.3375, + "tRecall_90": 0.1875, + "token_acc": 0.9273574561403509 + }, + { + "epoch": 0.185522610568163, + "grad_norm": 0.7195932865142822, + "learning_rate": 9.380188786670658e-06, + "loss": 0.1929931640625, + "step": 560, + "tIoU": 0.47695555686950686, + "tRecall_50": 0.49375, + "tRecall_75": 0.35, + "tRecall_90": 0.23125, + "token_acc": 0.9284353741496598 + }, + { + "epoch": 0.18717906244823587, + "grad_norm": 0.6872114539146423, + "learning_rate": 9.367190284545087e-06, + "loss": 0.2121337890625, + "step": 565, + "tIoU": 0.3961095571517944, + "tRecall_50": 0.4375, + "tRecall_75": 0.25625, + "tRecall_90": 0.13125, + "token_acc": 0.9261285909712722 + }, + { + "epoch": 0.18883551432830875, + "grad_norm": 0.6529171466827393, + "learning_rate": 9.354066093010898e-06, + "loss": 0.18876953125, + "step": 570, + "tIoU": 0.4421760082244873, + "tRecall_50": 0.475, + "tRecall_75": 0.3125, + "tRecall_90": 0.2125, + "token_acc": 0.9333150384193194 + }, + { + "epoch": 0.19049196620838166, + "grad_norm": 0.9455483555793762, + "learning_rate": 9.34081658978735e-06, + "loss": 0.2117919921875, + "step": 575, + "tIoU": 0.41131224632263186, + "tRecall_50": 0.4375, + "tRecall_75": 0.29375, + "tRecall_90": 0.1625, + "token_acc": 0.9229092890220648 + }, + { + "epoch": 0.19214841808845454, + "grad_norm": 0.6294817924499512, + "learning_rate": 9.327442156200217e-06, + "loss": 0.20238037109375, + "step": 580, + "tIoU": 0.42276864051818847, + "tRecall_50": 0.43125, + "tRecall_75": 0.2875, + "tRecall_90": 0.16875, + "token_acc": 0.9247135842880524 + }, + { + "epoch": 0.19380486996852742, + "grad_norm": 0.49901363253593445, + "learning_rate": 9.31394317717082e-06, + "loss": 0.19185791015625, + "step": 585, + "tIoU": 0.4232339859008789, + "tRecall_50": 0.46875, + "tRecall_75": 0.3125, + "tRecall_90": 0.1375, + "token_acc": 0.9246463547334058 + }, + { + "epoch": 0.1954613218486003, + "grad_norm": 0.6557704210281372, + "learning_rate": 9.300320041204947e-06, + "loss": 0.19619140625, + "step": 590, + "tIoU": 0.44829621315002444, + "tRecall_50": 0.5125, + "tRecall_75": 0.275, + "tRecall_90": 0.16875, + "token_acc": 0.9294503691550451 + }, + { + "epoch": 0.19711777372867317, + "grad_norm": 1.5239434242248535, + "learning_rate": 9.286573140381663e-06, + "loss": 0.2063232421875, + "step": 595, + "tIoU": 0.4620360851287842, + "tRecall_50": 0.5125, + "tRecall_75": 0.325, + "tRecall_90": 0.16875, + "token_acc": 0.926655719759168 + }, + { + "epoch": 0.19877422560874605, + "grad_norm": 0.68989497423172, + "learning_rate": 9.27270287034204e-06, + "loss": 0.20166015625, + "step": 600, + "tIoU": 0.44564409255981446, + "tRecall_50": 0.49375, + "tRecall_75": 0.2875, + "tRecall_90": 0.15625, + "token_acc": 0.9264705882352942 + }, + { + "epoch": 0.20043067748881896, + "grad_norm": 0.8628115057945251, + "learning_rate": 9.258709630277758e-06, + "loss": 0.19608154296875, + "step": 605, + "tIoU": 0.4197608470916748, + "tRecall_50": 0.425, + "tRecall_75": 0.29375, + "tRecall_90": 0.16875, + "token_acc": 0.930574462292404 + }, + { + "epoch": 0.20208712936889184, + "grad_norm": 0.5018081068992615, + "learning_rate": 9.244593822919627e-06, + "loss": 0.21024169921875, + "step": 610, + "tIoU": 0.4352898120880127, + "tRecall_50": 0.44375, + "tRecall_75": 0.275, + "tRecall_90": 0.15625, + "token_acc": 0.9244721169463995 + }, + { + "epoch": 0.20374358124896472, + "grad_norm": 0.5807973146438599, + "learning_rate": 9.230355854525988e-06, + "loss": 0.1876708984375, + "step": 615, + "tIoU": 0.45960350036621095, + "tRecall_50": 0.475, + "tRecall_75": 0.3125, + "tRecall_90": 0.19375, + "token_acc": 0.931185144729656 + }, + { + "epoch": 0.2054000331290376, + "grad_norm": 0.8211379647254944, + "learning_rate": 9.215996134871022e-06, + "loss": 0.2001220703125, + "step": 620, + "tIoU": 0.478743839263916, + "tRecall_50": 0.4875, + "tRecall_75": 0.325, + "tRecall_90": 0.16875, + "token_acc": 0.9288816503800217 + }, + { + "epoch": 0.20705648500911047, + "grad_norm": 0.7232769727706909, + "learning_rate": 9.201515077232958e-06, + "loss": 0.19661865234375, + "step": 625, + "tIoU": 0.4671457290649414, + "tRecall_50": 0.5, + "tRecall_75": 0.3375, + "tRecall_90": 0.16875, + "token_acc": 0.9303401360544218 + }, + { + "epoch": 0.20871293688918338, + "grad_norm": 0.7774529457092285, + "learning_rate": 9.186913098382185e-06, + "loss": 0.20330810546875, + "step": 630, + "tIoU": 0.446498441696167, + "tRecall_50": 0.46875, + "tRecall_75": 0.30625, + "tRecall_90": 0.15, + "token_acc": 0.924986361156574 + }, + { + "epoch": 0.21036938876925626, + "grad_norm": 0.5614256262779236, + "learning_rate": 9.172190618569236e-06, + "loss": 0.18858642578125, + "step": 635, + "tIoU": 0.4410062313079834, + "tRecall_50": 0.48125, + "tRecall_75": 0.29375, + "tRecall_90": 0.15625, + "token_acc": 0.9271232876712329 + }, + { + "epoch": 0.21202584064932914, + "grad_norm": 0.5532062649726868, + "learning_rate": 9.157348061512728e-06, + "loss": 0.19208984375, + "step": 640, + "tIoU": 0.4408290386199951, + "tRecall_50": 0.48125, + "tRecall_75": 0.2625, + "tRecall_90": 0.175, + "token_acc": 0.9247751430907605 + }, + { + "epoch": 0.21368229252940202, + "grad_norm": 0.6056038737297058, + "learning_rate": 9.142385854387128e-06, + "loss": 0.20494384765625, + "step": 645, + "tIoU": 0.4316275119781494, + "tRecall_50": 0.44375, + "tRecall_75": 0.2875, + "tRecall_90": 0.18125, + "token_acc": 0.9252464403066812 + }, + { + "epoch": 0.2153387444094749, + "grad_norm": 0.8482238054275513, + "learning_rate": 9.127304427810491e-06, + "loss": 0.18748779296875, + "step": 650, + "tIoU": 0.4653031349182129, + "tRecall_50": 0.53125, + "tRecall_75": 0.3, + "tRecall_90": 0.175, + "token_acc": 0.9333515432941819 + }, + { + "epoch": 0.21699519628954778, + "grad_norm": 0.543383777141571, + "learning_rate": 9.112104215832047e-06, + "loss": 0.1937255859375, + "step": 655, + "tIoU": 0.4971604824066162, + "tRecall_50": 0.5625, + "tRecall_75": 0.375, + "tRecall_90": 0.19375, + "token_acc": 0.9318366242899648 + }, + { + "epoch": 0.21865164816962068, + "grad_norm": 0.6806201338768005, + "learning_rate": 9.096785655919717e-06, + "loss": 0.192578125, + "step": 660, + "tIoU": 0.4546171188354492, + "tRecall_50": 0.4875, + "tRecall_75": 0.30625, + "tRecall_90": 0.1625, + "token_acc": 0.9255637055148057 + }, + { + "epoch": 0.22030810004969356, + "grad_norm": 0.5829448103904724, + "learning_rate": 9.08134918894752e-06, + "loss": 0.19219970703125, + "step": 665, + "tIoU": 0.4513428211212158, + "tRecall_50": 0.48125, + "tRecall_75": 0.29375, + "tRecall_90": 0.175, + "token_acc": 0.9289395172226743 + }, + { + "epoch": 0.22196455192976644, + "grad_norm": 0.8667960166931152, + "learning_rate": 9.06579525918289e-06, + "loss": 0.20467529296875, + "step": 670, + "tIoU": 0.42314701080322265, + "tRecall_50": 0.45, + "tRecall_75": 0.2625, + "tRecall_90": 0.15625, + "token_acc": 0.9248839114995903 + }, + { + "epoch": 0.22362100380983932, + "grad_norm": 0.720363974571228, + "learning_rate": 9.050124314273876e-06, + "loss": 0.192578125, + "step": 675, + "tIoU": 0.46733717918395995, + "tRecall_50": 0.4625, + "tRecall_75": 0.33125, + "tRecall_90": 0.2375, + "token_acc": 0.9315143246930423 + }, + { + "epoch": 0.2252774556899122, + "grad_norm": 0.6191056966781616, + "learning_rate": 9.034336805236279e-06, + "loss": 0.18389892578125, + "step": 680, + "tIoU": 0.48230271339416503, + "tRecall_50": 0.50625, + "tRecall_75": 0.34375, + "tRecall_90": 0.23125, + "token_acc": 0.9328950972336346 + }, + { + "epoch": 0.2269339075699851, + "grad_norm": 1.0231218338012695, + "learning_rate": 9.018433186440648e-06, + "loss": 0.18983154296875, + "step": 685, + "tIoU": 0.4781657695770264, + "tRecall_50": 0.5375, + "tRecall_75": 0.31875, + "tRecall_90": 0.16875, + "token_acc": 0.9283378746594005 + }, + { + "epoch": 0.22859035945005798, + "grad_norm": 0.6080626249313354, + "learning_rate": 9.002413915599223e-06, + "loss": 0.19080810546875, + "step": 690, + "tIoU": 0.42686448097229, + "tRecall_50": 0.45, + "tRecall_75": 0.2875, + "tRecall_90": 0.175, + "token_acc": 0.9304109589041096 + }, + { + "epoch": 0.23024681133013086, + "grad_norm": 0.6328234076499939, + "learning_rate": 8.986279453752758e-06, + "loss": 0.18778076171875, + "step": 695, + "tIoU": 0.46008977890014646, + "tRecall_50": 0.48125, + "tRecall_75": 0.36875, + "tRecall_90": 0.20625, + "token_acc": 0.9322080043561122 + }, + { + "epoch": 0.23190326321020374, + "grad_norm": 0.5854917764663696, + "learning_rate": 8.970030265257237e-06, + "loss": 0.19652099609375, + "step": 700, + "tIoU": 0.4481083393096924, + "tRecall_50": 0.48125, + "tRecall_75": 0.3125, + "tRecall_90": 0.1875, + "token_acc": 0.9253077975376197 + }, + { + "epoch": 0.23355971509027662, + "grad_norm": 0.563603401184082, + "learning_rate": 8.95366681777053e-06, + "loss": 0.184967041015625, + "step": 705, + "tIoU": 0.4827601909637451, + "tRecall_50": 0.5, + "tRecall_75": 0.3, + "tRecall_90": 0.175, + "token_acc": 0.9291575492341356 + }, + { + "epoch": 0.2352161669703495, + "grad_norm": 0.6367529034614563, + "learning_rate": 8.93718958223892e-06, + "loss": 0.178515625, + "step": 710, + "tIoU": 0.46496214866638186, + "tRecall_50": 0.5375, + "tRecall_75": 0.3, + "tRecall_90": 0.19375, + "token_acc": 0.930536638518115 + }, + { + "epoch": 0.2368726188504224, + "grad_norm": 0.5359927415847778, + "learning_rate": 8.920599032883553e-06, + "loss": 0.1771484375, + "step": 715, + "tIoU": 0.5005170345306397, + "tRecall_50": 0.55, + "tRecall_75": 0.33125, + "tRecall_90": 0.20625, + "token_acc": 0.9344620425996724 + }, + { + "epoch": 0.23852907073049529, + "grad_norm": 1.3301435708999634, + "learning_rate": 8.903895647186795e-06, + "loss": 0.18304443359375, + "step": 720, + "tIoU": 0.47890748977661135, + "tRecall_50": 0.525, + "tRecall_75": 0.3375, + "tRecall_90": 0.175, + "token_acc": 0.9330966684871654 + }, + { + "epoch": 0.24018552261056816, + "grad_norm": 0.9849122762680054, + "learning_rate": 8.887079905878484e-06, + "loss": 0.19580078125, + "step": 725, + "tIoU": 0.4430722713470459, + "tRecall_50": 0.4625, + "tRecall_75": 0.29375, + "tRecall_90": 0.20625, + "token_acc": 0.927891156462585 + }, + { + "epoch": 0.24184197449064104, + "grad_norm": 0.4319475293159485, + "learning_rate": 8.870152292922088e-06, + "loss": 0.17957763671875, + "step": 730, + "tIoU": 0.46705260276794436, + "tRecall_50": 0.5125, + "tRecall_75": 0.3125, + "tRecall_90": 0.19375, + "token_acc": 0.9331331877729258 + }, + { + "epoch": 0.24349842637071392, + "grad_norm": 0.7241905331611633, + "learning_rate": 8.853113295500795e-06, + "loss": 0.19373779296875, + "step": 735, + "tIoU": 0.4641860485076904, + "tRecall_50": 0.4875, + "tRecall_75": 0.29375, + "tRecall_90": 0.19375, + "token_acc": 0.9301440608861103 + }, + { + "epoch": 0.24515487825078683, + "grad_norm": 1.020323395729065, + "learning_rate": 8.83596340400347e-06, + "loss": 0.18074951171875, + "step": 740, + "tIoU": 0.4771578788757324, + "tRecall_50": 0.50625, + "tRecall_75": 0.38125, + "tRecall_90": 0.2625, + "token_acc": 0.934342069710889 + }, + { + "epoch": 0.2468113301308597, + "grad_norm": 0.6605811715126038, + "learning_rate": 8.818703112010562e-06, + "loss": 0.19556884765625, + "step": 745, + "tIoU": 0.4769937992095947, + "tRecall_50": 0.51875, + "tRecall_75": 0.3375, + "tRecall_90": 0.2125, + "token_acc": 0.9324104234527687 + }, + { + "epoch": 0.2484677820109326, + "grad_norm": 0.737177848815918, + "learning_rate": 8.801332916279882e-06, + "loss": 0.178759765625, + "step": 750, + "tIoU": 0.4775440216064453, + "tRecall_50": 0.4875, + "tRecall_75": 0.3375, + "tRecall_90": 0.2125, + "token_acc": 0.9352028314729104 + }, + { + "epoch": 0.25012423389100547, + "grad_norm": 1.0642855167388916, + "learning_rate": 8.783853316732313e-06, + "loss": 0.1890380859375, + "step": 755, + "tIoU": 0.48411898612976073, + "tRecall_50": 0.53125, + "tRecall_75": 0.375, + "tRecall_90": 0.20625, + "token_acc": 0.9313725490196079 + }, + { + "epoch": 0.25178068577107837, + "grad_norm": 0.8085300922393799, + "learning_rate": 8.766264816437424e-06, + "loss": 0.17894287109375, + "step": 760, + "tIoU": 0.4733723163604736, + "tRecall_50": 0.4875, + "tRecall_75": 0.36875, + "tRecall_90": 0.25, + "token_acc": 0.9372955288985824 + }, + { + "epoch": 0.2534371376511512, + "grad_norm": 1.0604671239852905, + "learning_rate": 8.74856792159899e-06, + "loss": 0.19639892578125, + "step": 765, + "tIoU": 0.4630789279937744, + "tRecall_50": 0.49375, + "tRecall_75": 0.325, + "tRecall_90": 0.175, + "token_acc": 0.9286300246103364 + }, + { + "epoch": 0.25509358953122413, + "grad_norm": 0.749838650226593, + "learning_rate": 8.730763141540417e-06, + "loss": 0.18775634765625, + "step": 770, + "tIoU": 0.46113018989562987, + "tRecall_50": 0.5125, + "tRecall_75": 0.31875, + "tRecall_90": 0.18125, + "token_acc": 0.9296004349007883 + }, + { + "epoch": 0.256750041411297, + "grad_norm": 0.7101616859436035, + "learning_rate": 8.712850988690094e-06, + "loss": 0.1763427734375, + "step": 775, + "tIoU": 0.4654840469360352, + "tRecall_50": 0.50625, + "tRecall_75": 0.34375, + "tRecall_90": 0.14375, + "token_acc": 0.9309309309309309 + }, + { + "epoch": 0.2584064932913699, + "grad_norm": 0.8688730001449585, + "learning_rate": 8.69483197856664e-06, + "loss": 0.18895263671875, + "step": 780, + "tIoU": 0.49425182342529295, + "tRecall_50": 0.51875, + "tRecall_75": 0.3375, + "tRecall_90": 0.2, + "token_acc": 0.9316146540027137 + }, + { + "epoch": 0.2600629451714428, + "grad_norm": 0.4564657211303711, + "learning_rate": 8.676706629764066e-06, + "loss": 0.18460693359375, + "step": 785, + "tIoU": 0.4841297149658203, + "tRecall_50": 0.48125, + "tRecall_75": 0.38125, + "tRecall_90": 0.24375, + "token_acc": 0.9336762317271251 + }, + { + "epoch": 0.26171939705151565, + "grad_norm": 0.6564610600471497, + "learning_rate": 8.658475463936849e-06, + "loss": 0.17938232421875, + "step": 790, + "tIoU": 0.4766863822937012, + "tRecall_50": 0.53125, + "tRecall_75": 0.3, + "tRecall_90": 0.20625, + "token_acc": 0.9309594998640935 + }, + { + "epoch": 0.26337584893158855, + "grad_norm": 0.587030827999115, + "learning_rate": 8.640139005784924e-06, + "loss": 0.1773681640625, + "step": 795, + "tIoU": 0.4590290069580078, + "tRecall_50": 0.49375, + "tRecall_75": 0.31875, + "tRecall_90": 0.21875, + "token_acc": 0.9338753387533876 + }, + { + "epoch": 0.2650323008116614, + "grad_norm": 0.594359815120697, + "learning_rate": 8.621697783038579e-06, + "loss": 0.19100341796875, + "step": 800, + "tIoU": 0.44837188720703125, + "tRecall_50": 0.48125, + "tRecall_75": 0.3375, + "tRecall_90": 0.175, + "token_acc": 0.927969556944822 + }, + { + "epoch": 0.2666887526917343, + "grad_norm": 1.1288412809371948, + "learning_rate": 8.603152326443262e-06, + "loss": 0.19122314453125, + "step": 805, + "tIoU": 0.44898138046264646, + "tRecall_50": 0.48125, + "tRecall_75": 0.2875, + "tRecall_90": 0.2, + "token_acc": 0.9313351498637602 + }, + { + "epoch": 0.2683452045718072, + "grad_norm": 0.6252028942108154, + "learning_rate": 8.584503169744317e-06, + "loss": 0.18260498046875, + "step": 810, + "tIoU": 0.4822195529937744, + "tRecall_50": 0.5125, + "tRecall_75": 0.3125, + "tRecall_90": 0.21875, + "token_acc": 0.9337513631406761 + }, + { + "epoch": 0.27000165645188007, + "grad_norm": 0.6308260560035706, + "learning_rate": 8.565750849671616e-06, + "loss": 0.17353515625, + "step": 815, + "tIoU": 0.5001877307891845, + "tRecall_50": 0.51875, + "tRecall_75": 0.38125, + "tRecall_90": 0.225, + "token_acc": 0.9359149168257431 + }, + { + "epoch": 0.271658108331953, + "grad_norm": 0.9611879587173462, + "learning_rate": 8.546895905924108e-06, + "loss": 0.1913330078125, + "step": 820, + "tIoU": 0.47109642028808596, + "tRecall_50": 0.4875, + "tRecall_75": 0.375, + "tRecall_90": 0.21875, + "token_acc": 0.9290780141843972 + }, + { + "epoch": 0.2733145602120258, + "grad_norm": 0.6250311732292175, + "learning_rate": 8.527938881154298e-06, + "loss": 0.1817138671875, + "step": 825, + "tIoU": 0.4798725128173828, + "tRecall_50": 0.53125, + "tRecall_75": 0.36875, + "tRecall_90": 0.1875, + "token_acc": 0.9346816069964471 + }, + { + "epoch": 0.27497101209209873, + "grad_norm": 0.4956880807876587, + "learning_rate": 8.508880320952614e-06, + "loss": 0.174755859375, + "step": 830, + "tIoU": 0.49635658264160154, + "tRecall_50": 0.5375, + "tRecall_75": 0.33125, + "tRecall_90": 0.21875, + "token_acc": 0.9361353711790393 + }, + { + "epoch": 0.2766274639721716, + "grad_norm": 1.0761417150497437, + "learning_rate": 8.489720773831717e-06, + "loss": 0.1855224609375, + "step": 835, + "tIoU": 0.44313597679138184, + "tRecall_50": 0.46875, + "tRecall_75": 0.2875, + "tRecall_90": 0.1875, + "token_acc": 0.9296276013143483 + }, + { + "epoch": 0.2782839158522445, + "grad_norm": 0.7068014740943909, + "learning_rate": 8.470460791210713e-06, + "loss": 0.19713134765625, + "step": 840, + "tIoU": 0.44738025665283204, + "tRecall_50": 0.48125, + "tRecall_75": 0.26875, + "tRecall_90": 0.14375, + "token_acc": 0.9266304347826086 + }, + { + "epoch": 0.2799403677323174, + "grad_norm": 0.5590986013412476, + "learning_rate": 8.451100927399272e-06, + "loss": 0.1806640625, + "step": 845, + "tIoU": 0.4611639499664307, + "tRecall_50": 0.51875, + "tRecall_75": 0.325, + "tRecall_90": 0.175, + "token_acc": 0.9299206131946346 + }, + { + "epoch": 0.28159681961239025, + "grad_norm": 0.7248173356056213, + "learning_rate": 8.431641739581688e-06, + "loss": 0.18641357421875, + "step": 850, + "tIoU": 0.4512800693511963, + "tRecall_50": 0.48125, + "tRecall_75": 0.30625, + "tRecall_90": 0.18125, + "token_acc": 0.9287849959228051 + }, + { + "epoch": 0.28325327149246315, + "grad_norm": 0.4190671443939209, + "learning_rate": 8.41208378780084e-06, + "loss": 0.194873046875, + "step": 855, + "tIoU": 0.4547574520111084, + "tRecall_50": 0.5, + "tRecall_75": 0.28125, + "tRecall_90": 0.19375, + "token_acc": 0.9293286219081273 + }, + { + "epoch": 0.284909723372536, + "grad_norm": 0.5794289708137512, + "learning_rate": 8.392427634942069e-06, + "loss": 0.179071044921875, + "step": 860, + "tIoU": 0.5311003684997558, + "tRecall_50": 0.58125, + "tRecall_75": 0.35, + "tRecall_90": 0.23125, + "token_acc": 0.9356948228882834 + }, + { + "epoch": 0.2865661752526089, + "grad_norm": 0.602472722530365, + "learning_rate": 8.372673846716977e-06, + "loss": 0.19573974609375, + "step": 865, + "tIoU": 0.4420654296875, + "tRecall_50": 0.46875, + "tRecall_75": 0.26875, + "tRecall_90": 0.175, + "token_acc": 0.9296619411123228 + }, + { + "epoch": 0.2882226271326818, + "grad_norm": 0.8868810534477234, + "learning_rate": 8.352822991647156e-06, + "loss": 0.20216064453125, + "step": 870, + "tIoU": 0.37321641445159914, + "tRecall_50": 0.39375, + "tRecall_75": 0.25, + "tRecall_90": 0.15, + "token_acc": 0.9270233196159122 + }, + { + "epoch": 0.28987907901275467, + "grad_norm": 0.5824978947639465, + "learning_rate": 8.332875641047817e-06, + "loss": 0.185546875, + "step": 875, + "tIoU": 0.42017297744750975, + "tRecall_50": 0.45625, + "tRecall_75": 0.2875, + "tRecall_90": 0.15, + "token_acc": 0.930733569675484 + }, + { + "epoch": 0.2915355308928276, + "grad_norm": 0.5793182849884033, + "learning_rate": 8.312832369011345e-06, + "loss": 0.18995361328125, + "step": 880, + "tIoU": 0.48091964721679686, + "tRecall_50": 0.49375, + "tRecall_75": 0.35625, + "tRecall_90": 0.25625, + "token_acc": 0.9309031556039173 + }, + { + "epoch": 0.29319198277290043, + "grad_norm": 0.44487687945365906, + "learning_rate": 8.292693752390787e-06, + "loss": 0.1759521484375, + "step": 885, + "tIoU": 0.508469533920288, + "tRecall_50": 0.54375, + "tRecall_75": 0.325, + "tRecall_90": 0.20625, + "token_acc": 0.9326738114866722 + }, + { + "epoch": 0.29484843465297333, + "grad_norm": 1.132308006286621, + "learning_rate": 8.272460370783238e-06, + "loss": 0.1759033203125, + "step": 890, + "tIoU": 0.44657158851623535, + "tRecall_50": 0.475, + "tRecall_75": 0.3, + "tRecall_90": 0.175, + "token_acc": 0.9313246471226927 + }, + { + "epoch": 0.29650488653304624, + "grad_norm": 0.5606683492660522, + "learning_rate": 8.25213280651317e-06, + "loss": 0.17841796875, + "step": 895, + "tIoU": 0.45934696197509767, + "tRecall_50": 0.50625, + "tRecall_75": 0.3, + "tRecall_90": 0.20625, + "token_acc": 0.9355537503384782 + }, + { + "epoch": 0.2981613384131191, + "grad_norm": 0.9272470474243164, + "learning_rate": 8.231711644615667e-06, + "loss": 0.19599609375, + "step": 900, + "tIoU": 0.3931166172027588, + "tRecall_50": 0.4125, + "tRecall_75": 0.2375, + "tRecall_90": 0.14375, + "token_acc": 0.9235374149659864 + }, + { + "epoch": 0.299817790293192, + "grad_norm": 0.586313784122467, + "learning_rate": 8.211197472819584e-06, + "loss": 0.18623046875, + "step": 905, + "tIoU": 0.46013169288635253, + "tRecall_50": 0.475, + "tRecall_75": 0.30625, + "tRecall_90": 0.19375, + "token_acc": 0.9299206131946346 + }, + { + "epoch": 0.30147424217326485, + "grad_norm": 0.7881156802177429, + "learning_rate": 8.190590881530644e-06, + "loss": 0.19263916015625, + "step": 910, + "tIoU": 0.4399313449859619, + "tRecall_50": 0.5125, + "tRecall_75": 0.3, + "tRecall_90": 0.15625, + "token_acc": 0.9282191780821918 + }, + { + "epoch": 0.30313069405333776, + "grad_norm": 0.6620801687240601, + "learning_rate": 8.169892463814435e-06, + "loss": 0.17945556640625, + "step": 915, + "tIoU": 0.5031002521514892, + "tRecall_50": 0.55, + "tRecall_75": 0.34375, + "tRecall_90": 0.20625, + "token_acc": 0.9306122448979591 + }, + { + "epoch": 0.3047871459334106, + "grad_norm": 0.5018839836120605, + "learning_rate": 8.149102815379341e-06, + "loss": 0.169970703125, + "step": 920, + "tIoU": 0.5297323703765869, + "tRecall_50": 0.58125, + "tRecall_75": 0.3875, + "tRecall_90": 0.225, + "token_acc": 0.9386252045826514 + }, + { + "epoch": 0.3064435978134835, + "grad_norm": 0.7223777174949646, + "learning_rate": 8.128222534559406e-06, + "loss": 0.19112548828125, + "step": 925, + "tIoU": 0.49451518058776855, + "tRecall_50": 0.525, + "tRecall_75": 0.35, + "tRecall_90": 0.225, + "token_acc": 0.9320467518347377 + }, + { + "epoch": 0.3081000496935564, + "grad_norm": 0.5758301615715027, + "learning_rate": 8.107252222297109e-06, + "loss": 0.184698486328125, + "step": 930, + "tIoU": 0.46448659896850586, + "tRecall_50": 0.5, + "tRecall_75": 0.325, + "tRecall_90": 0.2, + "token_acc": 0.9319672131147541 + }, + { + "epoch": 0.3097565015736293, + "grad_norm": 0.524838924407959, + "learning_rate": 8.086192482126063e-06, + "loss": 0.1820556640625, + "step": 935, + "tIoU": 0.4440475940704346, + "tRecall_50": 0.45625, + "tRecall_75": 0.3125, + "tRecall_90": 0.1625, + "token_acc": 0.9292324442025041 + }, + { + "epoch": 0.3114129534537022, + "grad_norm": 0.563029944896698, + "learning_rate": 8.065043920153658e-06, + "loss": 0.1949951171875, + "step": 940, + "tIoU": 0.4718454837799072, + "tRecall_50": 0.4875, + "tRecall_75": 0.31875, + "tRecall_90": 0.18125, + "token_acc": 0.9278518921862238 + }, + { + "epoch": 0.31306940533377503, + "grad_norm": 0.6538600325584412, + "learning_rate": 8.043807145043604e-06, + "loss": 0.19881591796875, + "step": 945, + "tIoU": 0.434765100479126, + "tRecall_50": 0.48125, + "tRecall_75": 0.3, + "tRecall_90": 0.1625, + "token_acc": 0.9281241491968418 + }, + { + "epoch": 0.31472585721384794, + "grad_norm": 0.47911351919174194, + "learning_rate": 8.02248276799842e-06, + "loss": 0.20108642578125, + "step": 950, + "tIoU": 0.436666202545166, + "tRecall_50": 0.4375, + "tRecall_75": 0.28125, + "tRecall_90": 0.18125, + "token_acc": 0.9261580381471389 + }, + { + "epoch": 0.31638230909392084, + "grad_norm": 0.8824852108955383, + "learning_rate": 8.001071402741843e-06, + "loss": 0.194287109375, + "step": 955, + "tIoU": 0.4746551990509033, + "tRecall_50": 0.5, + "tRecall_75": 0.35, + "tRecall_90": 0.19375, + "token_acc": 0.9284739284739285 + }, + { + "epoch": 0.3180387609739937, + "grad_norm": 1.3164430856704712, + "learning_rate": 7.979573665501161e-06, + "loss": 0.1749969482421875, + "step": 960, + "tIoU": 0.45885448455810546, + "tRecall_50": 0.46875, + "tRecall_75": 0.34375, + "tRecall_90": 0.2, + "token_acc": 0.9300603400987384 + }, + { + "epoch": 0.3196952128540666, + "grad_norm": 0.744648814201355, + "learning_rate": 7.957990174989487e-06, + "loss": 0.18673095703125, + "step": 965, + "tIoU": 0.4387523174285889, + "tRecall_50": 0.45, + "tRecall_75": 0.275, + "tRecall_90": 0.175, + "token_acc": 0.9319912948857454 + }, + { + "epoch": 0.32135166473413945, + "grad_norm": 0.4059516191482544, + "learning_rate": 7.936321552387942e-06, + "loss": 0.18834228515625, + "step": 970, + "tIoU": 0.4600379467010498, + "tRecall_50": 0.48125, + "tRecall_75": 0.31875, + "tRecall_90": 0.18125, + "token_acc": 0.9312789746386693 + }, + { + "epoch": 0.32300811661421236, + "grad_norm": 0.582944929599762, + "learning_rate": 7.914568421327782e-06, + "loss": 0.16092529296875, + "step": 975, + "tIoU": 0.5187951564788819, + "tRecall_50": 0.575, + "tRecall_75": 0.35625, + "tRecall_90": 0.2375, + "token_acc": 0.9401475813063679 + }, + { + "epoch": 0.32466456849428527, + "grad_norm": 0.5389725565910339, + "learning_rate": 7.892731407872446e-06, + "loss": 0.18634033203125, + "step": 980, + "tIoU": 0.45755715370178224, + "tRecall_50": 0.45625, + "tRecall_75": 0.31875, + "tRecall_90": 0.175, + "token_acc": 0.932322685435842 + }, + { + "epoch": 0.3263210203743581, + "grad_norm": 0.4865098297595978, + "learning_rate": 7.870811140499543e-06, + "loss": 0.17401123046875, + "step": 985, + "tIoU": 0.5109381675720215, + "tRecall_50": 0.5625, + "tRecall_75": 0.3625, + "tRecall_90": 0.2375, + "token_acc": 0.9346049046321526 + }, + { + "epoch": 0.327977472254431, + "grad_norm": 1.0831506252288818, + "learning_rate": 7.848808250082763e-06, + "loss": 0.16990966796875, + "step": 990, + "tIoU": 0.48476033210754393, + "tRecall_50": 0.5, + "tRecall_75": 0.36875, + "tRecall_90": 0.21875, + "token_acc": 0.9362865736942849 + }, + { + "epoch": 0.3296339241345039, + "grad_norm": 0.8786563277244568, + "learning_rate": 7.826723369873715e-06, + "loss": 0.18963623046875, + "step": 995, + "tIoU": 0.4693743705749512, + "tRecall_50": 0.5125, + "tRecall_75": 0.3875, + "tRecall_90": 0.21875, + "token_acc": 0.9316031359826981 + }, + { + "epoch": 0.3312903760145768, + "grad_norm": 0.5306078195571899, + "learning_rate": 7.80455713548371e-06, + "loss": 0.1913330078125, + "step": 1000, + "tIoU": 0.41149063110351564, + "tRecall_50": 0.44375, + "tRecall_75": 0.21875, + "tRecall_90": 0.13125, + "token_acc": 0.9273073781649878 + }, + { + "epoch": 0.33294682789464963, + "grad_norm": 0.5666636228561401, + "learning_rate": 7.782310184865463e-06, + "loss": 0.19432373046875, + "step": 1005, + "tIoU": 0.4761760711669922, + "tRecall_50": 0.51875, + "tRecall_75": 0.3375, + "tRecall_90": 0.19375, + "token_acc": 0.9221200648999459 + }, + { + "epoch": 0.33460327977472254, + "grad_norm": 0.6108797788619995, + "learning_rate": 7.759983158294727e-06, + "loss": 0.17784423828125, + "step": 1010, + "tIoU": 0.4799777984619141, + "tRecall_50": 0.5, + "tRecall_75": 0.35625, + "tRecall_90": 0.2125, + "token_acc": 0.931607629427793 + }, + { + "epoch": 0.33625973165479545, + "grad_norm": 0.540955126285553, + "learning_rate": 7.737576698351878e-06, + "loss": 0.18638916015625, + "step": 1015, + "tIoU": 0.4404765129089355, + "tRecall_50": 0.48125, + "tRecall_75": 0.26875, + "tRecall_90": 0.125, + "token_acc": 0.928024083196497 + }, + { + "epoch": 0.3379161835348683, + "grad_norm": 0.5134669542312622, + "learning_rate": 7.715091449903414e-06, + "loss": 0.1887939453125, + "step": 1020, + "tIoU": 0.4650413990020752, + "tRecall_50": 0.5, + "tRecall_75": 0.3, + "tRecall_90": 0.1375, + "token_acc": 0.9318986652138382 + }, + { + "epoch": 0.3395726354149412, + "grad_norm": 1.0283092260360718, + "learning_rate": 7.692528060083392e-06, + "loss": 0.18924560546875, + "step": 1025, + "tIoU": 0.4443990230560303, + "tRecall_50": 0.4625, + "tRecall_75": 0.28125, + "tRecall_90": 0.20625, + "token_acc": 0.9303849303849304 + }, + { + "epoch": 0.34122908729501406, + "grad_norm": 0.5505083799362183, + "learning_rate": 7.669887178274815e-06, + "loss": 0.17459716796875, + "step": 1030, + "tIoU": 0.5016412734985352, + "tRecall_50": 0.50625, + "tRecall_75": 0.425, + "tRecall_90": 0.275, + "token_acc": 0.9353260869565218 + }, + { + "epoch": 0.34288553917508696, + "grad_norm": 0.6912838816642761, + "learning_rate": 7.647169456090925e-06, + "loss": 0.177587890625, + "step": 1035, + "tIoU": 0.48855109214782716, + "tRecall_50": 0.56875, + "tRecall_75": 0.34375, + "tRecall_90": 0.21875, + "token_acc": 0.935685210312076 + }, + { + "epoch": 0.34454199105515987, + "grad_norm": 0.778060257434845, + "learning_rate": 7.624375547356473e-06, + "loss": 0.17293701171875, + "step": 1040, + "tIoU": 0.49210290908813475, + "tRecall_50": 0.50625, + "tRecall_75": 0.375, + "tRecall_90": 0.23125, + "token_acc": 0.936542669584245 + }, + { + "epoch": 0.3461984429352327, + "grad_norm": 0.9876009225845337, + "learning_rate": 7.601506108088874e-06, + "loss": 0.178436279296875, + "step": 1045, + "tIoU": 0.5314978122711181, + "tRecall_50": 0.575, + "tRecall_75": 0.3875, + "tRecall_90": 0.2375, + "token_acc": 0.9381639880141651 + }, + { + "epoch": 0.3478548948153056, + "grad_norm": 0.8202823400497437, + "learning_rate": 7.578561796479347e-06, + "loss": 0.18966064453125, + "step": 1050, + "tIoU": 0.4976532459259033, + "tRecall_50": 0.5375, + "tRecall_75": 0.325, + "tRecall_90": 0.175, + "token_acc": 0.9336076817558299 + }, + { + "epoch": 0.3495113466953785, + "grad_norm": 0.7027230858802795, + "learning_rate": 7.55554327287397e-06, + "loss": 0.19049072265625, + "step": 1055, + "tIoU": 0.4707040786743164, + "tRecall_50": 0.5, + "tRecall_75": 0.31875, + "tRecall_90": 0.16875, + "token_acc": 0.9283964062074599 + }, + { + "epoch": 0.3511677985754514, + "grad_norm": 0.5952333211898804, + "learning_rate": 7.532451199754662e-06, + "loss": 0.1842529296875, + "step": 1060, + "tIoU": 0.5045681476593018, + "tRecall_50": 0.55, + "tRecall_75": 0.3625, + "tRecall_90": 0.2125, + "token_acc": 0.9287849959228051 + }, + { + "epoch": 0.3528242504555243, + "grad_norm": 0.6814955472946167, + "learning_rate": 7.509286241720133e-06, + "loss": 0.178466796875, + "step": 1065, + "tIoU": 0.5122104167938233, + "tRecall_50": 0.5875, + "tRecall_75": 0.34375, + "tRecall_90": 0.19375, + "token_acc": 0.9327525183773482 + }, + { + "epoch": 0.35448070233559714, + "grad_norm": 0.7592156529426575, + "learning_rate": 7.486049065466743e-06, + "loss": 0.18367919921875, + "step": 1070, + "tIoU": 0.4885748863220215, + "tRecall_50": 0.525, + "tRecall_75": 0.34375, + "tRecall_90": 0.23125, + "token_acc": 0.935659760087241 + }, + { + "epoch": 0.35613715421567005, + "grad_norm": 0.6961227655410767, + "learning_rate": 7.462740339769323e-06, + "loss": 0.18165283203125, + "step": 1075, + "tIoU": 0.46857433319091796, + "tRecall_50": 0.5, + "tRecall_75": 0.3125, + "tRecall_90": 0.18125, + "token_acc": 0.9313725490196079 + }, + { + "epoch": 0.3577936060957429, + "grad_norm": 0.5508667230606079, + "learning_rate": 7.439360735461923e-06, + "loss": 0.18089599609375, + "step": 1080, + "tIoU": 0.48640642166137693, + "tRecall_50": 0.49375, + "tRecall_75": 0.36875, + "tRecall_90": 0.225, + "token_acc": 0.9371241115363587 + }, + { + "epoch": 0.3594500579758158, + "grad_norm": 0.5022640824317932, + "learning_rate": 7.415910925418508e-06, + "loss": 0.1744140625, + "step": 1085, + "tIoU": 0.5186983108520508, + "tRecall_50": 0.525, + "tRecall_75": 0.375, + "tRecall_90": 0.25, + "token_acc": 0.9383356070941337 + }, + { + "epoch": 0.3611065098558887, + "grad_norm": 0.5923255085945129, + "learning_rate": 7.392391584533591e-06, + "loss": 0.1818603515625, + "step": 1090, + "tIoU": 0.47256951332092284, + "tRecall_50": 0.51875, + "tRecall_75": 0.33125, + "tRecall_90": 0.175, + "token_acc": 0.9305177111716622 + }, + { + "epoch": 0.36276296173596156, + "grad_norm": 0.5832929611206055, + "learning_rate": 7.368803389702805e-06, + "loss": 0.1818603515625, + "step": 1095, + "tIoU": 0.49889507293701174, + "tRecall_50": 0.55625, + "tRecall_75": 0.325, + "tRecall_90": 0.2, + "token_acc": 0.9336986301369863 + }, + { + "epoch": 0.36441941361603447, + "grad_norm": 0.5508763790130615, + "learning_rate": 7.345147019803433e-06, + "loss": 0.1832275390625, + "step": 1100, + "tIoU": 0.47658720016479494, + "tRecall_50": 0.50625, + "tRecall_75": 0.3625, + "tRecall_90": 0.225, + "token_acc": 0.9340958605664488 + }, + { + "epoch": 0.3660758654961073, + "grad_norm": 0.49825698137283325, + "learning_rate": 7.321423155674858e-06, + "loss": 0.1625244140625, + "step": 1105, + "tIoU": 0.5512962341308594, + "tRecall_50": 0.5875, + "tRecall_75": 0.41875, + "tRecall_90": 0.26875, + "token_acc": 0.940515350877193 + }, + { + "epoch": 0.36773231737618023, + "grad_norm": 0.4958120286464691, + "learning_rate": 7.297632480098973e-06, + "loss": 0.1703369140625, + "step": 1110, + "tIoU": 0.529345703125, + "tRecall_50": 0.55, + "tRecall_75": 0.40625, + "tRecall_90": 0.25, + "token_acc": 0.9389084985066521 + }, + { + "epoch": 0.3693887692562531, + "grad_norm": 0.5963060259819031, + "learning_rate": 7.27377567778053e-06, + "loss": 0.17938232421875, + "step": 1115, + "tIoU": 0.5171398639678955, + "tRecall_50": 0.575, + "tRecall_75": 0.375, + "tRecall_90": 0.1875, + "token_acc": 0.9303523035230352 + }, + { + "epoch": 0.371045221136326, + "grad_norm": 0.7840315103530884, + "learning_rate": 7.249853435327435e-06, + "loss": 0.19072265625, + "step": 1120, + "tIoU": 0.45734376907348634, + "tRecall_50": 0.46875, + "tRecall_75": 0.29375, + "tRecall_90": 0.16875, + "token_acc": 0.9290375203915171 + }, + { + "epoch": 0.3727016730163989, + "grad_norm": 0.6449282169342041, + "learning_rate": 7.225866441230989e-06, + "loss": 0.18948974609375, + "step": 1125, + "tIoU": 0.4800254821777344, + "tRecall_50": 0.50625, + "tRecall_75": 0.325, + "tRecall_90": 0.2, + "token_acc": 0.9292709466811752 + }, + { + "epoch": 0.37435812489647174, + "grad_norm": 0.4980352520942688, + "learning_rate": 7.201815385846062e-06, + "loss": 0.17431640625, + "step": 1130, + "tIoU": 0.5075998783111573, + "tRecall_50": 0.53125, + "tRecall_75": 0.34375, + "tRecall_90": 0.25625, + "token_acc": 0.9373461012311901 + }, + { + "epoch": 0.37601457677654465, + "grad_norm": 1.3992658853530884, + "learning_rate": 7.177700961371239e-06, + "loss": 0.186663818359375, + "step": 1135, + "tIoU": 0.46782970428466797, + "tRecall_50": 0.475, + "tRecall_75": 0.3375, + "tRecall_90": 0.20625, + "token_acc": 0.9325751421608448 + }, + { + "epoch": 0.3776710286566175, + "grad_norm": 0.5823772549629211, + "learning_rate": 7.1535238618288915e-06, + "loss": 0.19114990234375, + "step": 1140, + "tIoU": 0.4538295269012451, + "tRecall_50": 0.46875, + "tRecall_75": 0.31875, + "tRecall_90": 0.20625, + "token_acc": 0.9287069666576308 + }, + { + "epoch": 0.3793274805366904, + "grad_norm": 0.5577098727226257, + "learning_rate": 7.1292847830452e-06, + "loss": 0.17542724609375, + "step": 1145, + "tIoU": 0.45658197402954104, + "tRecall_50": 0.49375, + "tRecall_75": 0.325, + "tRecall_90": 0.2125, + "token_acc": 0.9328621908127208 + }, + { + "epoch": 0.3809839324167633, + "grad_norm": 0.5321166515350342, + "learning_rate": 7.104984422630134e-06, + "loss": 0.1681640625, + "step": 1150, + "tIoU": 0.541117000579834, + "tRecall_50": 0.61875, + "tRecall_75": 0.35, + "tRecall_90": 0.2125, + "token_acc": 0.9354309165526675 + }, + { + "epoch": 0.38264038429683617, + "grad_norm": 0.5487158894538879, + "learning_rate": 7.080623479957372e-06, + "loss": 0.173046875, + "step": 1155, + "tIoU": 0.5375588893890381, + "tRecall_50": 0.5875, + "tRecall_75": 0.36875, + "tRecall_90": 0.25, + "token_acc": 0.937857726901063 + }, + { + "epoch": 0.3842968361769091, + "grad_norm": 0.640888512134552, + "learning_rate": 7.056202656144173e-06, + "loss": 0.18504638671875, + "step": 1160, + "tIoU": 0.4445330142974854, + "tRecall_50": 0.4625, + "tRecall_75": 0.28125, + "tRecall_90": 0.16875, + "token_acc": 0.930327868852459 + }, + { + "epoch": 0.3859532880569819, + "grad_norm": 0.47475653886795044, + "learning_rate": 7.031722654031192e-06, + "loss": 0.1691162109375, + "step": 1165, + "tIoU": 0.4704757213592529, + "tRecall_50": 0.50625, + "tRecall_75": 0.35, + "tRecall_90": 0.2, + "token_acc": 0.9352733206418276 + }, + { + "epoch": 0.38760973993705483, + "grad_norm": 0.7627384662628174, + "learning_rate": 7.007184178162268e-06, + "loss": 0.19022216796875, + "step": 1170, + "tIoU": 0.48237972259521483, + "tRecall_50": 0.51875, + "tRecall_75": 0.31875, + "tRecall_90": 0.2, + "token_acc": 0.9302832244008714 + }, + { + "epoch": 0.38926619181712774, + "grad_norm": 0.7071655988693237, + "learning_rate": 6.982587934764133e-06, + "loss": 0.17056884765625, + "step": 1175, + "tIoU": 0.5011544227600098, + "tRecall_50": 0.55, + "tRecall_75": 0.325, + "tRecall_90": 0.2, + "token_acc": 0.9356677524429967 + }, + { + "epoch": 0.3909226436972006, + "grad_norm": 0.6485012173652649, + "learning_rate": 6.9579346317260875e-06, + "loss": 0.1821044921875, + "step": 1180, + "tIoU": 0.5037876605987549, + "tRecall_50": 0.5625, + "tRecall_75": 0.3625, + "tRecall_90": 0.23125, + "token_acc": 0.9321340964840557 + }, + { + "epoch": 0.3925790955772735, + "grad_norm": 0.6403648853302002, + "learning_rate": 6.933224978579632e-06, + "loss": 0.167529296875, + "step": 1185, + "tIoU": 0.5066840648651123, + "tRecall_50": 0.55, + "tRecall_75": 0.3875, + "tRecall_90": 0.2125, + "token_acc": 0.938891906572515 + }, + { + "epoch": 0.39423554745734635, + "grad_norm": 0.5812480449676514, + "learning_rate": 6.90845968647805e-06, + "loss": 0.1739013671875, + "step": 1190, + "tIoU": 0.46291217803955076, + "tRecall_50": 0.525, + "tRecall_75": 0.33125, + "tRecall_90": 0.18125, + "token_acc": 0.9338415464198203 + }, + { + "epoch": 0.39589199933741925, + "grad_norm": 0.526985764503479, + "learning_rate": 6.883639468175926e-06, + "loss": 0.17183837890625, + "step": 1195, + "tIoU": 0.4848320484161377, + "tRecall_50": 0.55625, + "tRecall_75": 0.35, + "tRecall_90": 0.2, + "token_acc": 0.9363213992894234 + }, + { + "epoch": 0.3975484512174921, + "grad_norm": 0.7262851595878601, + "learning_rate": 6.85876503800865e-06, + "loss": 0.1871826171875, + "step": 1200, + "tIoU": 0.4523651599884033, + "tRecall_50": 0.4875, + "tRecall_75": 0.3125, + "tRecall_90": 0.2125, + "token_acc": 0.9293066088840737 + }, + { + "epoch": 0.399204903097565, + "grad_norm": 0.5683233141899109, + "learning_rate": 6.833837111871845e-06, + "loss": 0.19154052734375, + "step": 1205, + "tIoU": 0.43550262451171873, + "tRecall_50": 0.425, + "tRecall_75": 0.3125, + "tRecall_90": 0.21875, + "token_acc": 0.9319357473454941 + }, + { + "epoch": 0.4008613549776379, + "grad_norm": 0.703508734703064, + "learning_rate": 6.808856407200772e-06, + "loss": 0.177423095703125, + "step": 1210, + "tIoU": 0.525466012954712, + "tRecall_50": 0.58125, + "tRecall_75": 0.3625, + "tRecall_90": 0.2125, + "token_acc": 0.9351248642779587 + }, + { + "epoch": 0.40251780685771077, + "grad_norm": 0.5343628525733948, + "learning_rate": 6.7838236429496766e-06, + "loss": 0.16162109375, + "step": 1215, + "tIoU": 0.5424517631530762, + "tRecall_50": 0.55625, + "tRecall_75": 0.425, + "tRecall_90": 0.29375, + "token_acc": 0.9389645776566757 + }, + { + "epoch": 0.4041742587377837, + "grad_norm": 0.3772442936897278, + "learning_rate": 6.758739539571096e-06, + "loss": 0.1860107421875, + "step": 1220, + "tIoU": 0.4494819641113281, + "tRecall_50": 0.44375, + "tRecall_75": 0.3, + "tRecall_90": 0.19375, + "token_acc": 0.928922908693275 + }, + { + "epoch": 0.4058307106178565, + "grad_norm": 0.4595545530319214, + "learning_rate": 6.733604818995133e-06, + "loss": 0.17764892578125, + "step": 1225, + "tIoU": 0.5240142822265625, + "tRecall_50": 0.59375, + "tRecall_75": 0.3625, + "tRecall_90": 0.18125, + "token_acc": 0.9381639880141651 + }, + { + "epoch": 0.40748716249792943, + "grad_norm": 0.7019212245941162, + "learning_rate": 6.708420204608671e-06, + "loss": 0.1759033203125, + "step": 1230, + "tIoU": 0.43599753379821776, + "tRecall_50": 0.4625, + "tRecall_75": 0.2875, + "tRecall_90": 0.14375, + "token_acc": 0.9295238095238095 + }, + { + "epoch": 0.40914361437800234, + "grad_norm": 0.5394700169563293, + "learning_rate": 6.683186421234552e-06, + "loss": 0.153509521484375, + "step": 1235, + "tIoU": 0.5693307876586914, + "tRecall_50": 0.5875, + "tRecall_75": 0.43125, + "tRecall_90": 0.29375, + "token_acc": 0.9454693434617472 + }, + { + "epoch": 0.4108000662580752, + "grad_norm": 0.9691959619522095, + "learning_rate": 6.657904195110727e-06, + "loss": 0.182177734375, + "step": 1240, + "tIoU": 0.4553550720214844, + "tRecall_50": 0.48125, + "tRecall_75": 0.3, + "tRecall_90": 0.18125, + "token_acc": 0.9309120699071546 + }, + { + "epoch": 0.4124565181381481, + "grad_norm": 0.9197001457214355, + "learning_rate": 6.6325742538693424e-06, + "loss": 0.1696044921875, + "step": 1245, + "tIoU": 0.5319690227508544, + "tRecall_50": 0.5875, + "tRecall_75": 0.4, + "tRecall_90": 0.225, + "token_acc": 0.9379084967320261 + }, + { + "epoch": 0.41411297001822095, + "grad_norm": 0.7810382843017578, + "learning_rate": 6.607197326515808e-06, + "loss": 0.1841552734375, + "step": 1250, + "tIoU": 0.49736499786376953, + "tRecall_50": 0.5125, + "tRecall_75": 0.35, + "tRecall_90": 0.2625, + "token_acc": 0.9347237269772481 + }, + { + "epoch": 0.41576942189829386, + "grad_norm": 0.5846786499023438, + "learning_rate": 6.58177414340781e-06, + "loss": 0.1776611328125, + "step": 1255, + "tIoU": 0.495667028427124, + "tRecall_50": 0.50625, + "tRecall_75": 0.36875, + "tRecall_90": 0.25, + "token_acc": 0.9368852459016394 + }, + { + "epoch": 0.41742587377836676, + "grad_norm": 0.5197479724884033, + "learning_rate": 6.556305436234292e-06, + "loss": 0.17413330078125, + "step": 1260, + "tIoU": 0.5085415363311767, + "tRecall_50": 0.53125, + "tRecall_75": 0.36875, + "tRecall_90": 0.20625, + "token_acc": 0.9321477428180575 + }, + { + "epoch": 0.4190823256584396, + "grad_norm": 0.4828190207481384, + "learning_rate": 6.530791937994397e-06, + "loss": 0.1685791015625, + "step": 1265, + "tIoU": 0.557618522644043, + "tRecall_50": 0.61875, + "tRecall_75": 0.40625, + "tRecall_90": 0.25, + "token_acc": 0.9379254015790907 + }, + { + "epoch": 0.4207387775385125, + "grad_norm": 0.5395064949989319, + "learning_rate": 6.505234382976375e-06, + "loss": 0.18017578125, + "step": 1270, + "tIoU": 0.49621214866638186, + "tRecall_50": 0.54375, + "tRecall_75": 0.325, + "tRecall_90": 0.23125, + "token_acc": 0.935562805872757 + }, + { + "epoch": 0.42239522941858537, + "grad_norm": 0.6637852787971497, + "learning_rate": 6.479633506736447e-06, + "loss": 0.166064453125, + "step": 1275, + "tIoU": 0.4753843307495117, + "tRecall_50": 0.4875, + "tRecall_75": 0.38125, + "tRecall_90": 0.23125, + "token_acc": 0.9365600218758545 + }, + { + "epoch": 0.4240516812986583, + "grad_norm": 0.5723521113395691, + "learning_rate": 6.453990046077634e-06, + "loss": 0.15682373046875, + "step": 1280, + "tIoU": 0.5220492362976075, + "tRecall_50": 0.575, + "tRecall_75": 0.36875, + "tRecall_90": 0.2375, + "token_acc": 0.9406941787373599 + }, + { + "epoch": 0.4257081331787312, + "grad_norm": 0.5472649335861206, + "learning_rate": 6.42830473902855e-06, + "loss": 0.1759033203125, + "step": 1285, + "tIoU": 0.4785460472106934, + "tRecall_50": 0.5, + "tRecall_75": 0.3125, + "tRecall_90": 0.2375, + "token_acc": 0.937720968180582 + }, + { + "epoch": 0.42736458505880404, + "grad_norm": 0.5641521215438843, + "learning_rate": 6.402578324822169e-06, + "loss": 0.18299560546875, + "step": 1290, + "tIoU": 0.4435619354248047, + "tRecall_50": 0.4625, + "tRecall_75": 0.26875, + "tRecall_90": 0.18125, + "token_acc": 0.9317621266100301 + }, + { + "epoch": 0.42902103693887694, + "grad_norm": 0.4343439042568207, + "learning_rate": 6.3768115438745434e-06, + "loss": 0.184521484375, + "step": 1295, + "tIoU": 0.500169038772583, + "tRecall_50": 0.54375, + "tRecall_75": 0.3375, + "tRecall_90": 0.20625, + "token_acc": 0.9354926510615134 + }, + { + "epoch": 0.4306774888189498, + "grad_norm": 0.6340326070785522, + "learning_rate": 6.351005137763494e-06, + "loss": 0.1848388671875, + "step": 1300, + "tIoU": 0.513594675064087, + "tRecall_50": 0.5875, + "tRecall_75": 0.3375, + "tRecall_90": 0.175, + "token_acc": 0.9330430048992924 + }, + { + "epoch": 0.4323339406990227, + "grad_norm": 0.6288878917694092, + "learning_rate": 6.3251598492072685e-06, + "loss": 0.18204345703125, + "step": 1305, + "tIoU": 0.44141411781311035, + "tRecall_50": 0.45, + "tRecall_75": 0.28125, + "tRecall_90": 0.175, + "token_acc": 0.9306200491668943 + }, + { + "epoch": 0.43399039257909555, + "grad_norm": 0.7356960773468018, + "learning_rate": 6.299276422043171e-06, + "loss": 0.16866455078125, + "step": 1310, + "tIoU": 0.4666309833526611, + "tRecall_50": 0.525, + "tRecall_75": 0.35, + "tRecall_90": 0.23125, + "token_acc": 0.9368852459016394 + }, + { + "epoch": 0.43564684445916846, + "grad_norm": 0.6025828123092651, + "learning_rate": 6.273355601206143e-06, + "loss": 0.1773681640625, + "step": 1315, + "tIoU": 0.48667373657226565, + "tRecall_50": 0.54375, + "tRecall_75": 0.31875, + "tRecall_90": 0.2, + "token_acc": 0.9369171695008228 + }, + { + "epoch": 0.43730329633924137, + "grad_norm": 0.4284733831882477, + "learning_rate": 6.247398132707339e-06, + "loss": 0.180078125, + "step": 1320, + "tIoU": 0.49059696197509767, + "tRecall_50": 0.5125, + "tRecall_75": 0.36875, + "tRecall_90": 0.23125, + "token_acc": 0.9323737099402498 + }, + { + "epoch": 0.4389597482193142, + "grad_norm": 0.5862426161766052, + "learning_rate": 6.221404763612635e-06, + "loss": 0.18408203125, + "step": 1325, + "tIoU": 0.4608908653259277, + "tRecall_50": 0.5125, + "tRecall_75": 0.3, + "tRecall_90": 0.18125, + "token_acc": 0.9308365226899945 + }, + { + "epoch": 0.4406162000993871, + "grad_norm": 1.026079773902893, + "learning_rate": 6.1953762420211504e-06, + "loss": 0.1912109375, + "step": 1330, + "tIoU": 0.489334774017334, + "tRecall_50": 0.5125, + "tRecall_75": 0.3625, + "tRecall_90": 0.18125, + "token_acc": 0.9306122448979591 + }, + { + "epoch": 0.44227265197946, + "grad_norm": 0.6226865649223328, + "learning_rate": 6.169313317043702e-06, + "loss": 0.172412109375, + "step": 1335, + "tIoU": 0.482784366607666, + "tRecall_50": 0.525, + "tRecall_75": 0.31875, + "tRecall_90": 0.18125, + "token_acc": 0.9340809628008753 + }, + { + "epoch": 0.4439291038595329, + "grad_norm": 1.1509168148040771, + "learning_rate": 6.143216738781251e-06, + "loss": 0.1625244140625, + "step": 1340, + "tIoU": 0.5060029983520508, + "tRecall_50": 0.55625, + "tRecall_75": 0.35, + "tRecall_90": 0.2125, + "token_acc": 0.9401475813063679 + }, + { + "epoch": 0.4455855557396058, + "grad_norm": 0.6278771162033081, + "learning_rate": 6.117087258303314e-06, + "loss": 0.171435546875, + "step": 1345, + "tIoU": 0.4706563949584961, + "tRecall_50": 0.48125, + "tRecall_75": 0.375, + "tRecall_90": 0.21875, + "token_acc": 0.9342069342069342 + }, + { + "epoch": 0.44724200761967864, + "grad_norm": 0.4377037286758423, + "learning_rate": 6.09092562762634e-06, + "loss": 0.182470703125, + "step": 1350, + "tIoU": 0.43801226615905764, + "tRecall_50": 0.45, + "tRecall_75": 0.3375, + "tRecall_90": 0.2125, + "token_acc": 0.9311475409836065 + }, + { + "epoch": 0.44889845949975155, + "grad_norm": 0.8750120401382446, + "learning_rate": 6.064732599692079e-06, + "loss": 0.1802001953125, + "step": 1355, + "tIoU": 0.46314353942871095, + "tRecall_50": 0.50625, + "tRecall_75": 0.29375, + "tRecall_90": 0.20625, + "token_acc": 0.935025935025935 + }, + { + "epoch": 0.4505549113798244, + "grad_norm": 1.339943289756775, + "learning_rate": 6.038508928345906e-06, + "loss": 0.176611328125, + "step": 1360, + "tIoU": 0.46857519149780275, + "tRecall_50": 0.50625, + "tRecall_75": 0.31875, + "tRecall_90": 0.2, + "token_acc": 0.9331514324693042 + }, + { + "epoch": 0.4522113632598973, + "grad_norm": 0.38020825386047363, + "learning_rate": 6.012255368315123e-06, + "loss": 0.16986083984375, + "step": 1365, + "tIoU": 0.4757106781005859, + "tRecall_50": 0.5, + "tRecall_75": 0.35, + "tRecall_90": 0.18125, + "token_acc": 0.9373808875578546 + }, + { + "epoch": 0.4538678151399702, + "grad_norm": 0.5417476296424866, + "learning_rate": 5.985972675187239e-06, + "loss": 0.17435302734375, + "step": 1370, + "tIoU": 0.4798302173614502, + "tRecall_50": 0.5, + "tRecall_75": 0.30625, + "tRecall_90": 0.2, + "token_acc": 0.9314285714285714 + }, + { + "epoch": 0.45552426702004306, + "grad_norm": 0.6107211709022522, + "learning_rate": 5.959661605388229e-06, + "loss": 0.186907958984375, + "step": 1375, + "tIoU": 0.46599550247192384, + "tRecall_50": 0.45, + "tRecall_75": 0.3375, + "tRecall_90": 0.21875, + "token_acc": 0.930219929405376 + }, + { + "epoch": 0.45718071890011597, + "grad_norm": 1.218456506729126, + "learning_rate": 5.933322916160755e-06, + "loss": 0.163482666015625, + "step": 1380, + "tIoU": 0.5246377944946289, + "tRecall_50": 0.5625, + "tRecall_75": 0.375, + "tRecall_90": 0.23125, + "token_acc": 0.9383524277141299 + }, + { + "epoch": 0.4588371707801888, + "grad_norm": 0.46208956837654114, + "learning_rate": 5.906957365542376e-06, + "loss": 0.17071533203125, + "step": 1385, + "tIoU": 0.504871129989624, + "tRecall_50": 0.54375, + "tRecall_75": 0.3625, + "tRecall_90": 0.2125, + "token_acc": 0.9363686231486561 + }, + { + "epoch": 0.4604936226602617, + "grad_norm": 0.44012129306793213, + "learning_rate": 5.880565712343738e-06, + "loss": 0.18258056640625, + "step": 1390, + "tIoU": 0.49883341789245605, + "tRecall_50": 0.5375, + "tRecall_75": 0.35, + "tRecall_90": 0.19375, + "token_acc": 0.934135009565455 + }, + { + "epoch": 0.4621500745403346, + "grad_norm": 0.6960217356681824, + "learning_rate": 5.854148716126722e-06, + "loss": 0.16441650390625, + "step": 1395, + "tIoU": 0.5241262912750244, + "tRecall_50": 0.55625, + "tRecall_75": 0.3625, + "tRecall_90": 0.225, + "token_acc": 0.9384531590413944 + }, + { + "epoch": 0.4638065264204075, + "grad_norm": 1.0893362760543823, + "learning_rate": 5.827707137182597e-06, + "loss": 0.17698974609375, + "step": 1400, + "tIoU": 0.4990966796875, + "tRecall_50": 0.55, + "tRecall_75": 0.375, + "tRecall_90": 0.2, + "token_acc": 0.9318243796018544 + }, + { + "epoch": 0.4654629783004804, + "grad_norm": 0.6267316937446594, + "learning_rate": 5.801241736510128e-06, + "loss": 0.166265869140625, + "step": 1405, + "tIoU": 0.5334889411926269, + "tRecall_50": 0.55625, + "tRecall_75": 0.39375, + "tRecall_90": 0.25625, + "token_acc": 0.9387588459444747 + }, + { + "epoch": 0.46711943018055324, + "grad_norm": 0.5143325924873352, + "learning_rate": 5.774753275793682e-06, + "loss": 0.18515625, + "step": 1410, + "tIoU": 0.4607995986938477, + "tRecall_50": 0.5125, + "tRecall_75": 0.29375, + "tRecall_90": 0.175, + "token_acc": 0.930958904109589 + }, + { + "epoch": 0.46877588206062615, + "grad_norm": 0.4813113510608673, + "learning_rate": 5.748242517381301e-06, + "loss": 0.16951904296875, + "step": 1415, + "tIoU": 0.5051668643951416, + "tRecall_50": 0.56875, + "tRecall_75": 0.3625, + "tRecall_90": 0.225, + "token_acc": 0.9371086305472366 + }, + { + "epoch": 0.470432333940699, + "grad_norm": 0.7769686579704285, + "learning_rate": 5.721710224262759e-06, + "loss": 0.160137939453125, + "step": 1420, + "tIoU": 0.5131546974182128, + "tRecall_50": 0.5625, + "tRecall_75": 0.3875, + "tRecall_90": 0.24375, + "token_acc": 0.9359391965255157 + }, + { + "epoch": 0.4720887858207719, + "grad_norm": 0.5152138471603394, + "learning_rate": 5.695157160047618e-06, + "loss": 0.18056640625, + "step": 1425, + "tIoU": 0.4545309066772461, + "tRecall_50": 0.4875, + "tRecall_75": 0.325, + "tRecall_90": 0.23125, + "token_acc": 0.9355102040816327 + }, + { + "epoch": 0.4737452377008448, + "grad_norm": 0.6680763959884644, + "learning_rate": 5.66858408894323e-06, + "loss": 0.19742431640625, + "step": 1430, + "tIoU": 0.41425528526306155, + "tRecall_50": 0.43125, + "tRecall_75": 0.25, + "tRecall_90": 0.1625, + "token_acc": 0.9210740439381611 + }, + { + "epoch": 0.47540168958091766, + "grad_norm": 0.45692986249923706, + "learning_rate": 5.641991775732756e-06, + "loss": 0.16871337890625, + "step": 1435, + "tIoU": 0.5347061634063721, + "tRecall_50": 0.58125, + "tRecall_75": 0.4125, + "tRecall_90": 0.26875, + "token_acc": 0.9424069546318935 + }, + { + "epoch": 0.47705814146099057, + "grad_norm": 0.6588415503501892, + "learning_rate": 5.615380985753152e-06, + "loss": 0.15567626953125, + "step": 1440, + "tIoU": 0.5204037189483642, + "tRecall_50": 0.5625, + "tRecall_75": 0.34375, + "tRecall_90": 0.23125, + "token_acc": 0.935984745301008 + }, + { + "epoch": 0.4787145933410634, + "grad_norm": 0.6113463044166565, + "learning_rate": 5.588752484873146e-06, + "loss": 0.16971435546875, + "step": 1445, + "tIoU": 0.5141745567321777, + "tRecall_50": 0.55625, + "tRecall_75": 0.4, + "tRecall_90": 0.225, + "token_acc": 0.9385917030567685 + }, + { + "epoch": 0.48037104522113633, + "grad_norm": 0.45739471912384033, + "learning_rate": 5.562107039471187e-06, + "loss": 0.17086181640625, + "step": 1450, + "tIoU": 0.5260482788085937, + "tRecall_50": 0.5875, + "tRecall_75": 0.3875, + "tRecall_90": 0.2375, + "token_acc": 0.9395438307227261 + }, + { + "epoch": 0.48202749710120923, + "grad_norm": 0.510820209980011, + "learning_rate": 5.535445416413396e-06, + "loss": 0.17530517578125, + "step": 1455, + "tIoU": 0.51797456741333, + "tRecall_50": 0.55625, + "tRecall_75": 0.36875, + "tRecall_90": 0.2375, + "token_acc": 0.9394022484233616 + }, + { + "epoch": 0.4836839489812821, + "grad_norm": 0.6514920592308044, + "learning_rate": 5.508768383031497e-06, + "loss": 0.17376708984375, + "step": 1460, + "tIoU": 0.5032854080200195, + "tRecall_50": 0.55, + "tRecall_75": 0.38125, + "tRecall_90": 0.2125, + "token_acc": 0.9332247557003257 + }, + { + "epoch": 0.485340400861355, + "grad_norm": 0.5166162252426147, + "learning_rate": 5.482076707100723e-06, + "loss": 0.18035888671875, + "step": 1465, + "tIoU": 0.5216215133666993, + "tRecall_50": 0.5625, + "tRecall_75": 0.375, + "tRecall_90": 0.275, + "token_acc": 0.935230352303523 + }, + { + "epoch": 0.48699685274142784, + "grad_norm": 0.4673994183540344, + "learning_rate": 5.455371156817726e-06, + "loss": 0.16441650390625, + "step": 1470, + "tIoU": 0.5096771717071533, + "tRecall_50": 0.58125, + "tRecall_75": 0.34375, + "tRecall_90": 0.175, + "token_acc": 0.9382850901146914 + }, + { + "epoch": 0.48865330462150075, + "grad_norm": 0.5138466358184814, + "learning_rate": 5.428652500778472e-06, + "loss": 0.1756103515625, + "step": 1475, + "tIoU": 0.45171260833740234, + "tRecall_50": 0.45625, + "tRecall_75": 0.3, + "tRecall_90": 0.175, + "token_acc": 0.9321941958231624 + }, + { + "epoch": 0.49030975650157366, + "grad_norm": 0.3812209963798523, + "learning_rate": 5.401921507956113e-06, + "loss": 0.172747802734375, + "step": 1480, + "tIoU": 0.509559154510498, + "tRecall_50": 0.55, + "tRecall_75": 0.375, + "tRecall_90": 0.2375, + "token_acc": 0.9372111986952977 + }, + { + "epoch": 0.4919662083816465, + "grad_norm": 0.6065695285797119, + "learning_rate": 5.375178947678854e-06, + "loss": 0.188433837890625, + "step": 1485, + "tIoU": 0.44904017448425293, + "tRecall_50": 0.45625, + "tRecall_75": 0.3, + "tRecall_90": 0.18125, + "token_acc": 0.9303145336225597 + }, + { + "epoch": 0.4936226602617194, + "grad_norm": 0.9850072264671326, + "learning_rate": 5.348425589607823e-06, + "loss": 0.16478271484375, + "step": 1490, + "tIoU": 0.5062573909759521, + "tRecall_50": 0.54375, + "tRecall_75": 0.36875, + "tRecall_90": 0.24375, + "token_acc": 0.9359255202628697 + }, + { + "epoch": 0.49527911214179227, + "grad_norm": 0.7354980707168579, + "learning_rate": 5.321662203714909e-06, + "loss": 0.17548828125, + "step": 1495, + "tIoU": 0.5269069671630859, + "tRecall_50": 0.58125, + "tRecall_75": 0.40625, + "tRecall_90": 0.25, + "token_acc": 0.9357123399618632 + }, + { + "epoch": 0.4969355640218652, + "grad_norm": 0.4693378806114197, + "learning_rate": 5.294889560260605e-06, + "loss": 0.17823486328125, + "step": 1500, + "tIoU": 0.4679423809051514, + "tRecall_50": 0.50625, + "tRecall_75": 0.30625, + "tRecall_90": 0.19375, + "token_acc": 0.9336971350613915 + }, + { + "epoch": 0.498592015901938, + "grad_norm": 0.5371018648147583, + "learning_rate": 5.268108429771841e-06, + "loss": 0.166357421875, + "step": 1505, + "tIoU": 0.490740442276001, + "tRecall_50": 0.5625, + "tRecall_75": 0.35625, + "tRecall_90": 0.23125, + "token_acc": 0.9367675115835378 + }, + { + "epoch": 0.5002484677820109, + "grad_norm": 0.6271441578865051, + "learning_rate": 5.2413195830198075e-06, + "loss": 0.17530517578125, + "step": 1510, + "tIoU": 0.48148627281188966, + "tRecall_50": 0.5375, + "tRecall_75": 0.34375, + "tRecall_90": 0.2, + "token_acc": 0.9345026034530008 + }, + { + "epoch": 0.5019049196620838, + "grad_norm": 0.7134257555007935, + "learning_rate": 5.214523790997773e-06, + "loss": 0.1841796875, + "step": 1515, + "tIoU": 0.5060761451721192, + "tRecall_50": 0.5125, + "tRecall_75": 0.36875, + "tRecall_90": 0.24375, + "token_acc": 0.9378746594005449 + }, + { + "epoch": 0.5035613715421567, + "grad_norm": 0.5665849447250366, + "learning_rate": 5.187721824898889e-06, + "loss": 0.18895263671875, + "step": 1520, + "tIoU": 0.48246068954467775, + "tRecall_50": 0.50625, + "tRecall_75": 0.34375, + "tRecall_90": 0.19375, + "token_acc": 0.9330634278002699 + }, + { + "epoch": 0.5052178234222295, + "grad_norm": 0.7415385842323303, + "learning_rate": 5.160914456094005e-06, + "loss": 0.17012939453125, + "step": 1525, + "tIoU": 0.5014803409576416, + "tRecall_50": 0.5375, + "tRecall_75": 0.38125, + "tRecall_90": 0.20625, + "token_acc": 0.9352459016393443 + }, + { + "epoch": 0.5068742753023024, + "grad_norm": 0.5564184188842773, + "learning_rate": 5.134102456109461e-06, + "loss": 0.17379150390625, + "step": 1530, + "tIoU": 0.46170501708984374, + "tRecall_50": 0.48125, + "tRecall_75": 0.31875, + "tRecall_90": 0.19375, + "token_acc": 0.9353342428376534 + }, + { + "epoch": 0.5085307271823754, + "grad_norm": 0.5766486525535583, + "learning_rate": 5.10728659660488e-06, + "loss": 0.16895751953125, + "step": 1535, + "tIoU": 0.4611226558685303, + "tRecall_50": 0.5375, + "tRecall_75": 0.30625, + "tRecall_90": 0.1625, + "token_acc": 0.9382344902978956 + }, + { + "epoch": 0.5101871790624483, + "grad_norm": 0.48594367504119873, + "learning_rate": 5.080467649350968e-06, + "loss": 0.1735595703125, + "step": 1540, + "tIoU": 0.47720837593078613, + "tRecall_50": 0.51875, + "tRecall_75": 0.33125, + "tRecall_90": 0.19375, + "token_acc": 0.9352909189777052 + }, + { + "epoch": 0.5118436309425212, + "grad_norm": 0.47792771458625793, + "learning_rate": 5.053646386207296e-06, + "loss": 0.17591552734375, + "step": 1545, + "tIoU": 0.5278689861297607, + "tRecall_50": 0.60625, + "tRecall_75": 0.36875, + "tRecall_90": 0.20625, + "token_acc": 0.9347471451876019 + }, + { + "epoch": 0.513500082822594, + "grad_norm": 0.4938163757324219, + "learning_rate": 5.026823579100087e-06, + "loss": 0.1619140625, + "step": 1550, + "tIoU": 0.5453178882598877, + "tRecall_50": 0.575, + "tRecall_75": 0.4125, + "tRecall_90": 0.275, + "token_acc": 0.9415284199075333 + }, + { + "epoch": 0.5151565347026669, + "grad_norm": 0.7161099314689636, + "learning_rate": 5e-06, + "loss": 0.16890869140625, + "step": 1555, + "tIoU": 0.5169640064239502, + "tRecall_50": 0.54375, + "tRecall_75": 0.40625, + "tRecall_90": 0.25, + "token_acc": 0.9393939393939394 + }, + { + "epoch": 0.5168129865827398, + "grad_norm": 0.5965278148651123, + "learning_rate": 4.973176420899914e-06, + "loss": 0.16268310546875, + "step": 1560, + "tIoU": 0.5604077339172363, + "tRecall_50": 0.63125, + "tRecall_75": 0.40625, + "tRecall_90": 0.23125, + "token_acc": 0.9370932754880694 + }, + { + "epoch": 0.5184694384628127, + "grad_norm": 1.2650108337402344, + "learning_rate": 4.946353613792706e-06, + "loss": 0.17548828125, + "step": 1565, + "tIoU": 0.46562910079956055, + "tRecall_50": 0.50625, + "tRecall_75": 0.3625, + "tRecall_90": 0.19375, + "token_acc": 0.9333695947783519 + }, + { + "epoch": 0.5201258903428856, + "grad_norm": 0.45799070596694946, + "learning_rate": 4.919532350649033e-06, + "loss": 0.17598876953125, + "step": 1570, + "tIoU": 0.5043049335479737, + "tRecall_50": 0.55625, + "tRecall_75": 0.375, + "tRecall_90": 0.175, + "token_acc": 0.9324434758921275 + }, + { + "epoch": 0.5217823422229584, + "grad_norm": 0.6668441891670227, + "learning_rate": 4.8927134033951224e-06, + "loss": 0.17666015625, + "step": 1575, + "tIoU": 0.4904634475708008, + "tRecall_50": 0.50625, + "tRecall_75": 0.3875, + "tRecall_90": 0.225, + "token_acc": 0.9364130434782608 + }, + { + "epoch": 0.5234387941030313, + "grad_norm": 0.43300360441207886, + "learning_rate": 4.865897543890541e-06, + "loss": 0.155224609375, + "step": 1580, + "tIoU": 0.5382334232330322, + "tRecall_50": 0.5875, + "tRecall_75": 0.375, + "tRecall_90": 0.24375, + "token_acc": 0.9438970990695129 + }, + { + "epoch": 0.5250952459831042, + "grad_norm": 0.6325023174285889, + "learning_rate": 4.8390855439059955e-06, + "loss": 0.179248046875, + "step": 1585, + "tIoU": 0.46796865463256837, + "tRecall_50": 0.51875, + "tRecall_75": 0.2875, + "tRecall_90": 0.15, + "token_acc": 0.9322633297062024 + }, + { + "epoch": 0.5267516978631771, + "grad_norm": 0.49443551898002625, + "learning_rate": 4.812278175101112e-06, + "loss": 0.178515625, + "step": 1590, + "tIoU": 0.5053191661834717, + "tRecall_50": 0.50625, + "tRecall_75": 0.38125, + "tRecall_90": 0.25625, + "token_acc": 0.9367157665030006 + }, + { + "epoch": 0.52840814974325, + "grad_norm": 0.5553008317947388, + "learning_rate": 4.7854762090022274e-06, + "loss": 0.17664794921875, + "step": 1595, + "tIoU": 0.49562931060791016, + "tRecall_50": 0.55, + "tRecall_75": 0.34375, + "tRecall_90": 0.1875, + "token_acc": 0.9352557127312296 + }, + { + "epoch": 0.5300646016233228, + "grad_norm": 0.825242817401886, + "learning_rate": 4.758680416980193e-06, + "loss": 0.16912841796875, + "step": 1600, + "tIoU": 0.4986118793487549, + "tRecall_50": 0.55625, + "tRecall_75": 0.325, + "tRecall_90": 0.18125, + "token_acc": 0.9382984506659419 + }, + { + "epoch": 0.5317210535033957, + "grad_norm": 0.5749291777610779, + "learning_rate": 4.731891570228159e-06, + "loss": 0.175537109375, + "step": 1605, + "tIoU": 0.46056160926818845, + "tRecall_50": 0.50625, + "tRecall_75": 0.3625, + "tRecall_90": 0.18125, + "token_acc": 0.9333694474539544 + }, + { + "epoch": 0.5333775053834686, + "grad_norm": 0.5613634586334229, + "learning_rate": 4.7051104397393974e-06, + "loss": 0.1766845703125, + "step": 1610, + "tIoU": 0.4753167152404785, + "tRecall_50": 0.5, + "tRecall_75": 0.325, + "tRecall_90": 0.18125, + "token_acc": 0.9323737099402498 + }, + { + "epoch": 0.5350339572635415, + "grad_norm": 0.5255857110023499, + "learning_rate": 4.678337796285093e-06, + "loss": 0.17596435546875, + "step": 1615, + "tIoU": 0.5120919704437256, + "tRecall_50": 0.55625, + "tRecall_75": 0.3875, + "tRecall_90": 0.23125, + "token_acc": 0.9383692391600763 + }, + { + "epoch": 0.5366904091436144, + "grad_norm": 0.7099191546440125, + "learning_rate": 4.6515744103921786e-06, + "loss": 0.16297607421875, + "step": 1620, + "tIoU": 0.5286399841308593, + "tRecall_50": 0.58125, + "tRecall_75": 0.3625, + "tRecall_90": 0.25625, + "token_acc": 0.9428024083196497 + }, + { + "epoch": 0.5383468610236872, + "grad_norm": 0.5085911154747009, + "learning_rate": 4.6248210523211465e-06, + "loss": 0.16353759765625, + "step": 1625, + "tIoU": 0.5484652042388916, + "tRecall_50": 0.60625, + "tRecall_75": 0.36875, + "tRecall_90": 0.2125, + "token_acc": 0.9333151878062058 + }, + { + "epoch": 0.5400033129037601, + "grad_norm": 0.7192174792289734, + "learning_rate": 4.5980784920438886e-06, + "loss": 0.16826171875, + "step": 1630, + "tIoU": 0.5181742668151855, + "tRecall_50": 0.55, + "tRecall_75": 0.40625, + "tRecall_90": 0.21875, + "token_acc": 0.9386536373507057 + }, + { + "epoch": 0.541659764783833, + "grad_norm": 0.47222819924354553, + "learning_rate": 4.571347499221528e-06, + "loss": 0.1833984375, + "step": 1635, + "tIoU": 0.4572866916656494, + "tRecall_50": 0.4875, + "tRecall_75": 0.28125, + "tRecall_90": 0.175, + "token_acc": 0.9301250679717238 + }, + { + "epoch": 0.543316216663906, + "grad_norm": 0.6904683709144592, + "learning_rate": 4.544628843182274e-06, + "loss": 0.16632080078125, + "step": 1640, + "tIoU": 0.5224599838256836, + "tRecall_50": 0.56875, + "tRecall_75": 0.375, + "tRecall_90": 0.2375, + "token_acc": 0.9387755102040817 + }, + { + "epoch": 0.5449726685439787, + "grad_norm": 0.5094637274742126, + "learning_rate": 4.51792329289928e-06, + "loss": 0.17559814453125, + "step": 1645, + "tIoU": 0.5100062370300293, + "tRecall_50": 0.50625, + "tRecall_75": 0.36875, + "tRecall_90": 0.2375, + "token_acc": 0.9330966684871654 + }, + { + "epoch": 0.5466291204240517, + "grad_norm": 0.38745856285095215, + "learning_rate": 4.491231616968504e-06, + "loss": 0.168328857421875, + "step": 1650, + "tIoU": 0.4967671871185303, + "tRecall_50": 0.50625, + "tRecall_75": 0.35, + "tRecall_90": 0.2375, + "token_acc": 0.937636761487965 + }, + { + "epoch": 0.5482855723041246, + "grad_norm": 0.6616304516792297, + "learning_rate": 4.464554583586605e-06, + "loss": 0.157330322265625, + "step": 1655, + "tIoU": 0.5424611091613769, + "tRecall_50": 0.5875, + "tRecall_75": 0.3875, + "tRecall_90": 0.275, + "token_acc": 0.9404891304347827 + }, + { + "epoch": 0.5499420241841975, + "grad_norm": 3.821718215942383, + "learning_rate": 4.437892960528815e-06, + "loss": 0.18017578125, + "step": 1660, + "tIoU": 0.5014461040496826, + "tRecall_50": 0.5125, + "tRecall_75": 0.3625, + "tRecall_90": 0.25, + "token_acc": 0.9336235038084875 + }, + { + "epoch": 0.5515984760642704, + "grad_norm": 0.6461455225944519, + "learning_rate": 4.411247515126857e-06, + "loss": 0.16463623046875, + "step": 1665, + "tIoU": 0.4954252243041992, + "tRecall_50": 0.54375, + "tRecall_75": 0.3375, + "tRecall_90": 0.2, + "token_acc": 0.9375851730716817 + }, + { + "epoch": 0.5532549279443432, + "grad_norm": 0.3840186297893524, + "learning_rate": 4.384619014246848e-06, + "loss": 0.16251220703125, + "step": 1670, + "tIoU": 0.5271695613861084, + "tRecall_50": 0.55625, + "tRecall_75": 0.3875, + "tRecall_90": 0.25625, + "token_acc": 0.9403882964178288 + }, + { + "epoch": 0.5549113798244161, + "grad_norm": 0.6838501691818237, + "learning_rate": 4.358008224267245e-06, + "loss": 0.15831298828125, + "step": 1675, + "tIoU": 0.5436992168426513, + "tRecall_50": 0.5875, + "tRecall_75": 0.4375, + "tRecall_90": 0.2375, + "token_acc": 0.9403393541324576 + }, + { + "epoch": 0.556567831704489, + "grad_norm": 0.48298609256744385, + "learning_rate": 4.331415911056772e-06, + "loss": 0.17099609375, + "step": 1680, + "tIoU": 0.5099794864654541, + "tRecall_50": 0.525, + "tRecall_75": 0.4, + "tRecall_90": 0.23125, + "token_acc": 0.9344396082698585 + }, + { + "epoch": 0.5582242835845619, + "grad_norm": 0.6135856509208679, + "learning_rate": 4.304842839952384e-06, + "loss": 0.18450927734375, + "step": 1685, + "tIoU": 0.4841867446899414, + "tRecall_50": 0.55, + "tRecall_75": 0.3625, + "tRecall_90": 0.16875, + "token_acc": 0.9314658689148763 + }, + { + "epoch": 0.5598807354646348, + "grad_norm": 0.5720863342285156, + "learning_rate": 4.2782897757372414e-06, + "loss": 0.16536865234375, + "step": 1690, + "tIoU": 0.5474737644195556, + "tRecall_50": 0.6, + "tRecall_75": 0.40625, + "tRecall_90": 0.25, + "token_acc": 0.9392778993435449 + }, + { + "epoch": 0.5615371873447076, + "grad_norm": 0.5420379638671875, + "learning_rate": 4.251757482618701e-06, + "loss": 0.17178955078125, + "step": 1695, + "tIoU": 0.5092611312866211, + "tRecall_50": 0.5625, + "tRecall_75": 0.3625, + "tRecall_90": 0.24375, + "token_acc": 0.9342465753424658 + }, + { + "epoch": 0.5631936392247805, + "grad_norm": 0.6336923837661743, + "learning_rate": 4.22524672420632e-06, + "loss": 0.1845458984375, + "step": 1700, + "tIoU": 0.4917005062103271, + "tRecall_50": 0.55625, + "tRecall_75": 0.3875, + "tRecall_90": 0.2125, + "token_acc": 0.9361702127659575 + }, + { + "epoch": 0.5648500911048534, + "grad_norm": 0.9051311612129211, + "learning_rate": 4.1987582634898724e-06, + "loss": 0.18272705078125, + "step": 1705, + "tIoU": 0.5123229026794434, + "tRecall_50": 0.58125, + "tRecall_75": 0.36875, + "tRecall_90": 0.20625, + "token_acc": 0.9338820301783265 + }, + { + "epoch": 0.5665065429849263, + "grad_norm": 0.7399612665176392, + "learning_rate": 4.172292862817403e-06, + "loss": 0.160467529296875, + "step": 1710, + "tIoU": 0.5589471340179444, + "tRecall_50": 0.61875, + "tRecall_75": 0.39375, + "tRecall_90": 0.25, + "token_acc": 0.9445506692160612 + }, + { + "epoch": 0.5681629948649992, + "grad_norm": 0.41440925002098083, + "learning_rate": 4.14585128387328e-06, + "loss": 0.18409423828125, + "step": 1715, + "tIoU": 0.4675239086151123, + "tRecall_50": 0.525, + "tRecall_75": 0.35, + "tRecall_90": 0.19375, + "token_acc": 0.9282218597063622 + }, + { + "epoch": 0.569819446745072, + "grad_norm": 0.47274792194366455, + "learning_rate": 4.119434287656265e-06, + "loss": 0.167669677734375, + "step": 1720, + "tIoU": 0.5223277091979981, + "tRecall_50": 0.5875, + "tRecall_75": 0.35, + "tRecall_90": 0.19375, + "token_acc": 0.9380796508456083 + }, + { + "epoch": 0.5714758986251449, + "grad_norm": 0.49913594126701355, + "learning_rate": 4.0930426344576255e-06, + "loss": 0.162109375, + "step": 1725, + "tIoU": 0.45806708335876467, + "tRecall_50": 0.50625, + "tRecall_75": 0.325, + "tRecall_90": 0.21875, + "token_acc": 0.9379929290182214 + }, + { + "epoch": 0.5731323505052178, + "grad_norm": 1.1948215961456299, + "learning_rate": 4.0666770838392465e-06, + "loss": 0.18353271484375, + "step": 1730, + "tIoU": 0.5288844585418702, + "tRecall_50": 0.5625, + "tRecall_75": 0.38125, + "tRecall_90": 0.25, + "token_acc": 0.9384363933533098 + }, + { + "epoch": 0.5747888023852907, + "grad_norm": 0.4646267890930176, + "learning_rate": 4.040338394611772e-06, + "loss": 0.1573974609375, + "step": 1735, + "tIoU": 0.5586710453033448, + "tRecall_50": 0.6125, + "tRecall_75": 0.4125, + "tRecall_90": 0.275, + "token_acc": 0.941079295154185 + }, + { + "epoch": 0.5764452542653636, + "grad_norm": 0.5275194048881531, + "learning_rate": 4.014027324812761e-06, + "loss": 0.16785888671875, + "step": 1740, + "tIoU": 0.5046965122222901, + "tRecall_50": 0.5625, + "tRecall_75": 0.38125, + "tRecall_90": 0.20625, + "token_acc": 0.9382682327232996 + }, + { + "epoch": 0.5781017061454364, + "grad_norm": 0.5200470685958862, + "learning_rate": 3.987744631684877e-06, + "loss": 0.1590087890625, + "step": 1745, + "tIoU": 0.5480869293212891, + "tRecall_50": 0.63125, + "tRecall_75": 0.38125, + "tRecall_90": 0.25, + "token_acc": 0.9403919433859553 + }, + { + "epoch": 0.5797581580255093, + "grad_norm": 0.5766518712043762, + "learning_rate": 3.961491071654095e-06, + "loss": 0.174365234375, + "step": 1750, + "tIoU": 0.508606243133545, + "tRecall_50": 0.53125, + "tRecall_75": 0.3375, + "tRecall_90": 0.2375, + "token_acc": 0.9362224039247752 + }, + { + "epoch": 0.5814146099055822, + "grad_norm": 0.5985773801803589, + "learning_rate": 3.9352674003079225e-06, + "loss": 0.15159912109375, + "step": 1755, + "tIoU": 0.572104549407959, + "tRecall_50": 0.65, + "tRecall_75": 0.40625, + "tRecall_90": 0.2625, + "token_acc": 0.9412889131622064 + }, + { + "epoch": 0.5830710617856552, + "grad_norm": 0.9872517585754395, + "learning_rate": 3.909074372373662e-06, + "loss": 0.16317138671875, + "step": 1760, + "tIoU": 0.5388112545013428, + "tRecall_50": 0.58125, + "tRecall_75": 0.425, + "tRecall_90": 0.29375, + "token_acc": 0.94379581862612 + }, + { + "epoch": 0.5847275136657281, + "grad_norm": 0.5140849947929382, + "learning_rate": 3.882912741696688e-06, + "loss": 0.14564208984375, + "step": 1765, + "tIoU": 0.5485616683959961, + "tRecall_50": 0.58125, + "tRecall_75": 0.44375, + "tRecall_90": 0.28125, + "token_acc": 0.9443079443079443 + }, + { + "epoch": 0.5863839655458009, + "grad_norm": 0.6263390779495239, + "learning_rate": 3.85678326121875e-06, + "loss": 0.17509765625, + "step": 1770, + "tIoU": 0.5056493282318115, + "tRecall_50": 0.54375, + "tRecall_75": 0.4, + "tRecall_90": 0.21875, + "token_acc": 0.9353518821603928 + }, + { + "epoch": 0.5880404174258738, + "grad_norm": 0.6849157214164734, + "learning_rate": 3.830686682956299e-06, + "loss": 0.17532958984375, + "step": 1775, + "tIoU": 0.5150390148162842, + "tRecall_50": 0.5875, + "tRecall_75": 0.36875, + "tRecall_90": 0.2125, + "token_acc": 0.9324952172724789 + }, + { + "epoch": 0.5896968693059467, + "grad_norm": 0.8499537110328674, + "learning_rate": 3.804623757978849e-06, + "loss": 0.14559326171875, + "step": 1780, + "tIoU": 0.5719325542449951, + "tRecall_50": 0.625, + "tRecall_75": 0.425, + "tRecall_90": 0.325, + "token_acc": 0.9457406402604449 + }, + { + "epoch": 0.5913533211860196, + "grad_norm": 0.5121057033538818, + "learning_rate": 3.778595236387367e-06, + "loss": 0.1645263671875, + "step": 1785, + "tIoU": 0.5143845081329346, + "tRecall_50": 0.5625, + "tRecall_75": 0.40625, + "tRecall_90": 0.23125, + "token_acc": 0.9362224039247752 + }, + { + "epoch": 0.5930097730660925, + "grad_norm": 0.7534790635108948, + "learning_rate": 3.752601867292664e-06, + "loss": 0.17528076171875, + "step": 1790, + "tIoU": 0.4931938648223877, + "tRecall_50": 0.5375, + "tRecall_75": 0.31875, + "tRecall_90": 0.16875, + "token_acc": 0.9328256731030732 + }, + { + "epoch": 0.5946662249461653, + "grad_norm": 0.5419090986251831, + "learning_rate": 3.726644398793857e-06, + "loss": 0.16246337890625, + "step": 1795, + "tIoU": 0.5456307888031006, + "tRecall_50": 0.6375, + "tRecall_75": 0.3625, + "tRecall_90": 0.225, + "token_acc": 0.9386536373507057 + }, + { + "epoch": 0.5963226768262382, + "grad_norm": 0.5578802824020386, + "learning_rate": 3.7007235779568303e-06, + "loss": 0.18109130859375, + "step": 1800, + "tIoU": 0.4708672046661377, + "tRecall_50": 0.51875, + "tRecall_75": 0.29375, + "tRecall_90": 0.14375, + "token_acc": 0.9316146540027137 + }, + { + "epoch": 0.5979791287063111, + "grad_norm": 1.2091337442398071, + "learning_rate": 3.6748401507927327e-06, + "loss": 0.169842529296875, + "step": 1805, + "tIoU": 0.49648056030273435, + "tRecall_50": 0.525, + "tRecall_75": 0.35625, + "tRecall_90": 0.2125, + "token_acc": 0.9373644251626898 + }, + { + "epoch": 0.599635580586384, + "grad_norm": 0.6194023489952087, + "learning_rate": 3.648994862236508e-06, + "loss": 0.16412353515625, + "step": 1810, + "tIoU": 0.5116681098937989, + "tRecall_50": 0.58125, + "tRecall_75": 0.35625, + "tRecall_90": 0.175, + "token_acc": 0.9343724364232978 + }, + { + "epoch": 0.6012920324664568, + "grad_norm": 0.5854647159576416, + "learning_rate": 3.6231884561254582e-06, + "loss": 0.1682373046875, + "step": 1815, + "tIoU": 0.516246223449707, + "tRecall_50": 0.58125, + "tRecall_75": 0.35, + "tRecall_90": 0.21875, + "token_acc": 0.9384531590413944 + }, + { + "epoch": 0.6029484843465297, + "grad_norm": 0.7405495643615723, + "learning_rate": 3.5974216751778325e-06, + "loss": 0.1697265625, + "step": 1820, + "tIoU": 0.5363904476165772, + "tRecall_50": 0.58125, + "tRecall_75": 0.4125, + "tRecall_90": 0.21875, + "token_acc": 0.9363784665579119 + }, + { + "epoch": 0.6046049362266026, + "grad_norm": 0.45874157547950745, + "learning_rate": 3.5716952609714517e-06, + "loss": 0.1662109375, + "step": 1825, + "tIoU": 0.5525872707366943, + "tRecall_50": 0.61875, + "tRecall_75": 0.38125, + "tRecall_90": 0.2125, + "token_acc": 0.9399188092016239 + }, + { + "epoch": 0.6062613881066755, + "grad_norm": 0.6839113235473633, + "learning_rate": 3.546009953922368e-06, + "loss": 0.164111328125, + "step": 1830, + "tIoU": 0.49982576370239257, + "tRecall_50": 0.51875, + "tRecall_75": 0.325, + "tRecall_90": 0.20625, + "token_acc": 0.9340419732897247 + }, + { + "epoch": 0.6079178399867484, + "grad_norm": 0.6627931594848633, + "learning_rate": 3.520366493263554e-06, + "loss": 0.1625, + "step": 1835, + "tIoU": 0.5280773162841796, + "tRecall_50": 0.575, + "tRecall_75": 0.375, + "tRecall_90": 0.2375, + "token_acc": 0.9365641165260006 + }, + { + "epoch": 0.6095742918668212, + "grad_norm": 0.8185617327690125, + "learning_rate": 3.4947656170236255e-06, + "loss": 0.1634521484375, + "step": 1840, + "tIoU": 0.4996767520904541, + "tRecall_50": 0.5375, + "tRecall_75": 0.34375, + "tRecall_90": 0.2, + "token_acc": 0.9390710382513662 + }, + { + "epoch": 0.6112307437468941, + "grad_norm": 0.4964151382446289, + "learning_rate": 3.4692080620056044e-06, + "loss": 0.1591796875, + "step": 1845, + "tIoU": 0.5419831275939941, + "tRecall_50": 0.58125, + "tRecall_75": 0.40625, + "tRecall_90": 0.28125, + "token_acc": 0.941464742717125 + }, + { + "epoch": 0.612887195626967, + "grad_norm": 0.7755962610244751, + "learning_rate": 3.4436945637657106e-06, + "loss": 0.18828125, + "step": 1850, + "tIoU": 0.4821001052856445, + "tRecall_50": 0.5125, + "tRecall_75": 0.36875, + "tRecall_90": 0.175, + "token_acc": 0.9328785811732606 + }, + { + "epoch": 0.6145436475070399, + "grad_norm": 0.42528417706489563, + "learning_rate": 3.4182258565921933e-06, + "loss": 0.1717041015625, + "step": 1855, + "tIoU": 0.5007997512817383, + "tRecall_50": 0.5125, + "tRecall_75": 0.33125, + "tRecall_90": 0.19375, + "token_acc": 0.9337658826709921 + }, + { + "epoch": 0.6162000993871128, + "grad_norm": 0.549651563167572, + "learning_rate": 3.3928026734841935e-06, + "loss": 0.15889892578125, + "step": 1860, + "tIoU": 0.5112274646759033, + "tRecall_50": 0.54375, + "tRecall_75": 0.375, + "tRecall_90": 0.24375, + "token_acc": 0.9406641262928688 + }, + { + "epoch": 0.6178565512671856, + "grad_norm": 0.5841197967529297, + "learning_rate": 3.367425746130658e-06, + "loss": 0.167706298828125, + "step": 1865, + "tIoU": 0.510018253326416, + "tRecall_50": 0.55, + "tRecall_75": 0.3375, + "tRecall_90": 0.2375, + "token_acc": 0.9364949577541565 + }, + { + "epoch": 0.6195130031472585, + "grad_norm": 0.8597052097320557, + "learning_rate": 3.342095804889275e-06, + "loss": 0.16390380859375, + "step": 1870, + "tIoU": 0.5023562431335449, + "tRecall_50": 0.5625, + "tRecall_75": 0.3625, + "tRecall_90": 0.2, + "token_acc": 0.9369713506139155 + }, + { + "epoch": 0.6211694550273315, + "grad_norm": 0.4235498011112213, + "learning_rate": 3.3168135787654493e-06, + "loss": 0.17318115234375, + "step": 1875, + "tIoU": 0.4791114807128906, + "tRecall_50": 0.5375, + "tRecall_75": 0.3125, + "tRecall_90": 0.15625, + "token_acc": 0.934092758340114 + }, + { + "epoch": 0.6228259069074044, + "grad_norm": 0.5873208045959473, + "learning_rate": 3.291579795391331e-06, + "loss": 0.1798095703125, + "step": 1880, + "tIoU": 0.5017872333526612, + "tRecall_50": 0.55, + "tRecall_75": 0.375, + "tRecall_90": 0.2375, + "token_acc": 0.9341496598639456 + }, + { + "epoch": 0.6244823587874773, + "grad_norm": 0.27464815974235535, + "learning_rate": 3.2663951810048683e-06, + "loss": 0.16212158203125, + "step": 1885, + "tIoU": 0.5321149826049805, + "tRecall_50": 0.58125, + "tRecall_75": 0.3875, + "tRecall_90": 0.25625, + "token_acc": 0.9363162228932199 + }, + { + "epoch": 0.6261388106675501, + "grad_norm": 0.5738742351531982, + "learning_rate": 3.2412604604289073e-06, + "loss": 0.162310791015625, + "step": 1890, + "tIoU": 0.5201348781585693, + "tRecall_50": 0.56875, + "tRecall_75": 0.3875, + "tRecall_90": 0.23125, + "token_acc": 0.9401360544217687 + }, + { + "epoch": 0.627795262547623, + "grad_norm": 0.44134360551834106, + "learning_rate": 3.216176357050327e-06, + "loss": 0.181396484375, + "step": 1895, + "tIoU": 0.4499537467956543, + "tRecall_50": 0.4625, + "tRecall_75": 0.325, + "tRecall_90": 0.2, + "token_acc": 0.9313432835820895 + }, + { + "epoch": 0.6294517144276959, + "grad_norm": 0.49762654304504395, + "learning_rate": 3.1911435927992298e-06, + "loss": 0.1660888671875, + "step": 1900, + "tIoU": 0.5242837905883789, + "tRecall_50": 0.54375, + "tRecall_75": 0.375, + "tRecall_90": 0.25625, + "token_acc": 0.9374828626268166 + }, + { + "epoch": 0.6311081663077688, + "grad_norm": 0.6275957822799683, + "learning_rate": 3.1661628881281564e-06, + "loss": 0.156964111328125, + "step": 1905, + "tIoU": 0.562921667098999, + "tRecall_50": 0.59375, + "tRecall_75": 0.46875, + "tRecall_90": 0.3, + "token_acc": 0.9451119606772256 + }, + { + "epoch": 0.6327646181878417, + "grad_norm": 0.5966311693191528, + "learning_rate": 3.1412349619913514e-06, + "loss": 0.17091064453125, + "step": 1910, + "tIoU": 0.5571235179901123, + "tRecall_50": 0.6, + "tRecall_75": 0.45, + "tRecall_90": 0.2625, + "token_acc": 0.9354223433242507 + }, + { + "epoch": 0.6344210700679145, + "grad_norm": 0.4650852084159851, + "learning_rate": 3.116360531824074e-06, + "loss": 0.16427001953125, + "step": 1915, + "tIoU": 0.5082977294921875, + "tRecall_50": 0.55625, + "tRecall_75": 0.34375, + "tRecall_90": 0.2125, + "token_acc": 0.9358449358449359 + }, + { + "epoch": 0.6360775219479874, + "grad_norm": 0.534633457660675, + "learning_rate": 3.091540313521952e-06, + "loss": 0.16798095703125, + "step": 1920, + "tIoU": 0.493639612197876, + "tRecall_50": 0.525, + "tRecall_75": 0.375, + "tRecall_90": 0.26875, + "token_acc": 0.9388145315487572 + }, + { + "epoch": 0.6377339738280603, + "grad_norm": 0.3949797749519348, + "learning_rate": 3.0667750214203697e-06, + "loss": 0.1730224609375, + "step": 1925, + "tIoU": 0.5319153308868408, + "tRecall_50": 0.58125, + "tRecall_75": 0.40625, + "tRecall_90": 0.2625, + "token_acc": 0.9356328082563824 + }, + { + "epoch": 0.6393904257081332, + "grad_norm": 0.7183004021644592, + "learning_rate": 3.0420653682739154e-06, + "loss": 0.18037109375, + "step": 1930, + "tIoU": 0.4745781421661377, + "tRecall_50": 0.51875, + "tRecall_75": 0.3, + "tRecall_90": 0.16875, + "token_acc": 0.9310910582444627 + }, + { + "epoch": 0.6410468775882061, + "grad_norm": 0.4463639557361603, + "learning_rate": 3.0174120652358683e-06, + "loss": 0.16229248046875, + "step": 1935, + "tIoU": 0.52823486328125, + "tRecall_50": 0.55625, + "tRecall_75": 0.36875, + "tRecall_90": 0.24375, + "token_acc": 0.9381302807304442 + }, + { + "epoch": 0.6427033294682789, + "grad_norm": 0.812966525554657, + "learning_rate": 2.992815821837732e-06, + "loss": 0.171685791015625, + "step": 1940, + "tIoU": 0.504635238647461, + "tRecall_50": 0.5375, + "tRecall_75": 0.375, + "tRecall_90": 0.25625, + "token_acc": 0.9380771319934817 + }, + { + "epoch": 0.6443597813483518, + "grad_norm": 0.5824111104011536, + "learning_rate": 2.9682773459688087e-06, + "loss": 0.163983154296875, + "step": 1945, + "tIoU": 0.5575588703155517, + "tRecall_50": 0.60625, + "tRecall_75": 0.40625, + "tRecall_90": 0.25625, + "token_acc": 0.9411442649876813 + }, + { + "epoch": 0.6460162332284247, + "grad_norm": 0.754898190498352, + "learning_rate": 2.943797343855829e-06, + "loss": 0.15794677734375, + "step": 1950, + "tIoU": 0.5148393154144287, + "tRecall_50": 0.55, + "tRecall_75": 0.425, + "tRecall_90": 0.23125, + "token_acc": 0.9404567699836868 + }, + { + "epoch": 0.6476726851084976, + "grad_norm": 0.549222469329834, + "learning_rate": 2.919376520042628e-06, + "loss": 0.16236572265625, + "step": 1955, + "tIoU": 0.5246635913848877, + "tRecall_50": 0.575, + "tRecall_75": 0.4, + "tRecall_90": 0.20625, + "token_acc": 0.9405860010851872 + }, + { + "epoch": 0.6493291369885705, + "grad_norm": 0.5977454781532288, + "learning_rate": 2.895015577369866e-06, + "loss": 0.15792236328125, + "step": 1960, + "tIoU": 0.5148863315582275, + "tRecall_50": 0.56875, + "tRecall_75": 0.3875, + "tRecall_90": 0.23125, + "token_acc": 0.9405214557305812 + }, + { + "epoch": 0.6509855888686433, + "grad_norm": 0.40176132321357727, + "learning_rate": 2.8707152169548014e-06, + "loss": 0.15865478515625, + "step": 1965, + "tIoU": 0.5105422973632813, + "tRecall_50": 0.5375, + "tRecall_75": 0.35625, + "tRecall_90": 0.2625, + "token_acc": 0.9393527332064183 + }, + { + "epoch": 0.6526420407487162, + "grad_norm": 0.6446389555931091, + "learning_rate": 2.846476138171111e-06, + "loss": 0.15938720703125, + "step": 1970, + "tIoU": 0.5290540218353271, + "tRecall_50": 0.5625, + "tRecall_75": 0.36875, + "tRecall_90": 0.275, + "token_acc": 0.9428882240957303 + }, + { + "epoch": 0.6542984926287891, + "grad_norm": 0.41462159156799316, + "learning_rate": 2.822299038628762e-06, + "loss": 0.169287109375, + "step": 1975, + "tIoU": 0.5018656253814697, + "tRecall_50": 0.4875, + "tRecall_75": 0.35625, + "tRecall_90": 0.25625, + "token_acc": 0.9404081632653061 + }, + { + "epoch": 0.655954944508862, + "grad_norm": 0.9564310908317566, + "learning_rate": 2.7981846141539402e-06, + "loss": 0.16522216796875, + "step": 1980, + "tIoU": 0.5413733005523682, + "tRecall_50": 0.6, + "tRecall_75": 0.39375, + "tRecall_90": 0.2625, + "token_acc": 0.9384363933533098 + }, + { + "epoch": 0.657611396388935, + "grad_norm": 0.600674569606781, + "learning_rate": 2.774133558769014e-06, + "loss": 0.16549072265625, + "step": 1985, + "tIoU": 0.5427555561065673, + "tRecall_50": 0.6125, + "tRecall_75": 0.3875, + "tRecall_90": 0.275, + "token_acc": 0.9395336225596529 + }, + { + "epoch": 0.6592678482690078, + "grad_norm": 0.6821838021278381, + "learning_rate": 2.7501465646725646e-06, + "loss": 0.156494140625, + "step": 1990, + "tIoU": 0.5400360584259033, + "tRecall_50": 0.5625, + "tRecall_75": 0.4125, + "tRecall_90": 0.28125, + "token_acc": 0.9448689956331878 + }, + { + "epoch": 0.6609243001490807, + "grad_norm": 0.5166508555412292, + "learning_rate": 2.7262243222194728e-06, + "loss": 0.15616455078125, + "step": 1995, + "tIoU": 0.535555648803711, + "tRecall_50": 0.60625, + "tRecall_75": 0.36875, + "tRecall_90": 0.21875, + "token_acc": 0.9406130268199234 + }, + { + "epoch": 0.6625807520291536, + "grad_norm": 0.7035804390907288, + "learning_rate": 2.70236751990103e-06, + "loss": 0.17166748046875, + "step": 2000, + "tIoU": 0.5096180438995361, + "tRecall_50": 0.54375, + "tRecall_75": 0.3625, + "tRecall_90": 0.20625, + "token_acc": 0.9358172423171064 + }, + { + "epoch": 0.6642372039092265, + "grad_norm": 0.5203202962875366, + "learning_rate": 2.6785768443251437e-06, + "loss": 0.161578369140625, + "step": 2005, + "tIoU": 0.5287741184234619, + "tRecall_50": 0.5625, + "tRecall_75": 0.35625, + "tRecall_90": 0.23125, + "token_acc": 0.9400934322616104 + }, + { + "epoch": 0.6658936557892993, + "grad_norm": 0.8393142819404602, + "learning_rate": 2.6548529801965683e-06, + "loss": 0.168927001953125, + "step": 2010, + "tIoU": 0.5189241886138916, + "tRecall_50": 0.55625, + "tRecall_75": 0.40625, + "tRecall_90": 0.25, + "token_acc": 0.9400328587075575 + }, + { + "epoch": 0.6675501076693722, + "grad_norm": 0.4761408865451813, + "learning_rate": 2.6311966102971964e-06, + "loss": 0.16002197265625, + "step": 2015, + "tIoU": 0.49315404891967773, + "tRecall_50": 0.55, + "tRecall_75": 0.35625, + "tRecall_90": 0.2625, + "token_acc": 0.9369197160021846 + }, + { + "epoch": 0.6692065595494451, + "grad_norm": 0.638813316822052, + "learning_rate": 2.60760841546641e-06, + "loss": 0.163818359375, + "step": 2020, + "tIoU": 0.5471360683441162, + "tRecall_50": 0.5875, + "tRecall_75": 0.4, + "tRecall_90": 0.26875, + "token_acc": 0.9425006780580417 + }, + { + "epoch": 0.670863011429518, + "grad_norm": 0.47899043560028076, + "learning_rate": 2.5840890745814927e-06, + "loss": 0.17288818359375, + "step": 2025, + "tIoU": 0.49503450393676757, + "tRecall_50": 0.53125, + "tRecall_75": 0.33125, + "tRecall_90": 0.2625, + "token_acc": 0.9383019383019383 + }, + { + "epoch": 0.6725194633095909, + "grad_norm": 1.0909929275512695, + "learning_rate": 2.5606392645380783e-06, + "loss": 0.16365966796875, + "step": 2030, + "tIoU": 0.5265616416931153, + "tRecall_50": 0.56875, + "tRecall_75": 0.4, + "tRecall_90": 0.25, + "token_acc": 0.9414980863860033 + }, + { + "epoch": 0.6741759151896637, + "grad_norm": 0.57685786485672, + "learning_rate": 2.537259660230679e-06, + "loss": 0.15804443359375, + "step": 2035, + "tIoU": 0.5010174751281739, + "tRecall_50": 0.53125, + "tRecall_75": 0.4, + "tRecall_90": 0.23125, + "token_acc": 0.9364776444929117 + }, + { + "epoch": 0.6758323670697366, + "grad_norm": 0.8344061970710754, + "learning_rate": 2.513950934533259e-06, + "loss": 0.1663818359375, + "step": 2040, + "tIoU": 0.5199985027313232, + "tRecall_50": 0.50625, + "tRecall_75": 0.40625, + "tRecall_90": 0.28125, + "token_acc": 0.9410801963993454 + }, + { + "epoch": 0.6774888189498095, + "grad_norm": 0.7436922192573547, + "learning_rate": 2.4907137582798676e-06, + "loss": 0.1641357421875, + "step": 2045, + "tIoU": 0.4792342185974121, + "tRecall_50": 0.5125, + "tRecall_75": 0.3375, + "tRecall_90": 0.225, + "token_acc": 0.9375851730716817 + }, + { + "epoch": 0.6791452708298824, + "grad_norm": 0.5653188824653625, + "learning_rate": 2.4675488002453385e-06, + "loss": 0.1564208984375, + "step": 2050, + "tIoU": 0.5584266185760498, + "tRecall_50": 0.6, + "tRecall_75": 0.4625, + "tRecall_90": 0.2875, + "token_acc": 0.9416098226466576 + }, + { + "epoch": 0.6808017227099553, + "grad_norm": 0.5735906362533569, + "learning_rate": 2.4444567271260316e-06, + "loss": 0.17310791015625, + "step": 2055, + "tIoU": 0.49056329727172854, + "tRecall_50": 0.54375, + "tRecall_75": 0.3625, + "tRecall_90": 0.225, + "token_acc": 0.9352654387865655 + }, + { + "epoch": 0.6824581745900281, + "grad_norm": 0.734527051448822, + "learning_rate": 2.4214382035206525e-06, + "loss": 0.1667236328125, + "step": 2060, + "tIoU": 0.5110460758209229, + "tRecall_50": 0.56875, + "tRecall_75": 0.35, + "tRecall_90": 0.2, + "token_acc": 0.9353695118625579 + }, + { + "epoch": 0.684114626470101, + "grad_norm": 1.2782374620437622, + "learning_rate": 2.398493891911127e-06, + "loss": 0.15985107421875, + "step": 2065, + "tIoU": 0.564049768447876, + "tRecall_50": 0.60625, + "tRecall_75": 0.44375, + "tRecall_90": 0.2625, + "token_acc": 0.9417555373256767 + }, + { + "epoch": 0.6857710783501739, + "grad_norm": 0.5806080102920532, + "learning_rate": 2.3756244526435303e-06, + "loss": 0.164990234375, + "step": 2070, + "tIoU": 0.49380931854248045, + "tRecall_50": 0.54375, + "tRecall_75": 0.3625, + "tRecall_90": 0.24375, + "token_acc": 0.9363091997822537 + }, + { + "epoch": 0.6874275302302468, + "grad_norm": 0.599310040473938, + "learning_rate": 2.3528305439090744e-06, + "loss": 0.166650390625, + "step": 2075, + "tIoU": 0.5209931373596192, + "tRecall_50": 0.575, + "tRecall_75": 0.38125, + "tRecall_90": 0.2375, + "token_acc": 0.9398047722342733 + }, + { + "epoch": 0.6890839821103197, + "grad_norm": 0.6360598802566528, + "learning_rate": 2.3301128217251866e-06, + "loss": 0.16470947265625, + "step": 2080, + "tIoU": 0.5022995948791504, + "tRecall_50": 0.50625, + "tRecall_75": 0.36875, + "tRecall_90": 0.26875, + "token_acc": 0.9365773646801531 + }, + { + "epoch": 0.6907404339903925, + "grad_norm": 0.6225826144218445, + "learning_rate": 2.3074719399166085e-06, + "loss": 0.166156005859375, + "step": 2085, + "tIoU": 0.532801342010498, + "tRecall_50": 0.55625, + "tRecall_75": 0.4, + "tRecall_90": 0.26875, + "token_acc": 0.9410962639760022 + }, + { + "epoch": 0.6923968858704654, + "grad_norm": 0.543380081653595, + "learning_rate": 2.284908550096586e-06, + "loss": 0.1636474609375, + "step": 2090, + "tIoU": 0.49683642387390137, + "tRecall_50": 0.5375, + "tRecall_75": 0.3625, + "tRecall_90": 0.225, + "token_acc": 0.9385489983757445 + }, + { + "epoch": 0.6940533377505383, + "grad_norm": 0.6770683526992798, + "learning_rate": 2.2624233016481224e-06, + "loss": 0.168560791015625, + "step": 2095, + "tIoU": 0.4816957950592041, + "tRecall_50": 0.5125, + "tRecall_75": 0.35625, + "tRecall_90": 0.2375, + "token_acc": 0.9354223433242507 + }, + { + "epoch": 0.6957097896306113, + "grad_norm": 0.49540141224861145, + "learning_rate": 2.240016841705274e-06, + "loss": 0.1667724609375, + "step": 2100, + "tIoU": 0.5249676704406738, + "tRecall_50": 0.55625, + "tRecall_75": 0.4, + "tRecall_90": 0.23125, + "token_acc": 0.940424374319913 + }, + { + "epoch": 0.6973662415106842, + "grad_norm": 0.5058975219726562, + "learning_rate": 2.217689815134539e-06, + "loss": 0.1658203125, + "step": 2105, + "tIoU": 0.5202480792999268, + "tRecall_50": 0.54375, + "tRecall_75": 0.38125, + "tRecall_90": 0.25, + "token_acc": 0.9400054540496319 + }, + { + "epoch": 0.699022693390757, + "grad_norm": 0.3797245919704437, + "learning_rate": 2.1954428645162914e-06, + "loss": 0.164404296875, + "step": 2110, + "tIoU": 0.5144438743591309, + "tRecall_50": 0.54375, + "tRecall_75": 0.4125, + "tRecall_90": 0.2625, + "token_acc": 0.9379084967320261 + }, + { + "epoch": 0.7006791452708299, + "grad_norm": 0.6059390306472778, + "learning_rate": 2.173276630126287e-06, + "loss": 0.1622802734375, + "step": 2115, + "tIoU": 0.5094360828399658, + "tRecall_50": 0.5625, + "tRecall_75": 0.3625, + "tRecall_90": 0.225, + "token_acc": 0.9365986394557824 + }, + { + "epoch": 0.7023355971509028, + "grad_norm": 0.7819738984107971, + "learning_rate": 2.151191749917239e-06, + "loss": 0.16075439453125, + "step": 2120, + "tIoU": 0.5810178279876709, + "tRecall_50": 0.6125, + "tRecall_75": 0.43125, + "tRecall_90": 0.3125, + "token_acc": 0.9424754634678298 + }, + { + "epoch": 0.7039920490309757, + "grad_norm": 0.5492057800292969, + "learning_rate": 2.129188859500459e-06, + "loss": 0.16397705078125, + "step": 2125, + "tIoU": 0.5358535289764405, + "tRecall_50": 0.575, + "tRecall_75": 0.39375, + "tRecall_90": 0.225, + "token_acc": 0.9394104803493449 + }, + { + "epoch": 0.7056485009110486, + "grad_norm": 0.5017256736755371, + "learning_rate": 2.107268592127557e-06, + "loss": 0.163037109375, + "step": 2130, + "tIoU": 0.5214012145996094, + "tRecall_50": 0.60625, + "tRecall_75": 0.31875, + "tRecall_90": 0.2, + "token_acc": 0.9364949577541565 + }, + { + "epoch": 0.7073049527911214, + "grad_norm": 0.6092507839202881, + "learning_rate": 2.085431578672219e-06, + "loss": 0.17164306640625, + "step": 2135, + "tIoU": 0.5037929058074951, + "tRecall_50": 0.58125, + "tRecall_75": 0.33125, + "tRecall_90": 0.2, + "token_acc": 0.9367502726281353 + }, + { + "epoch": 0.7089614046711943, + "grad_norm": 0.5345575213432312, + "learning_rate": 2.0636784476120598e-06, + "loss": 0.16177978515625, + "step": 2140, + "tIoU": 0.5068227291107178, + "tRecall_50": 0.55, + "tRecall_75": 0.39375, + "tRecall_90": 0.20625, + "token_acc": 0.940424374319913 + }, + { + "epoch": 0.7106178565512672, + "grad_norm": 0.8379113674163818, + "learning_rate": 2.042009825010513e-06, + "loss": 0.161236572265625, + "step": 2145, + "tIoU": 0.5136396884918213, + "tRecall_50": 0.51875, + "tRecall_75": 0.35625, + "tRecall_90": 0.2375, + "token_acc": 0.9387087986924544 + }, + { + "epoch": 0.7122743084313401, + "grad_norm": 0.5098614692687988, + "learning_rate": 2.0204263344988395e-06, + "loss": 0.1553955078125, + "step": 2150, + "tIoU": 0.573584508895874, + "tRecall_50": 0.6375, + "tRecall_75": 0.4125, + "tRecall_90": 0.3, + "token_acc": 0.941320293398533 + }, + { + "epoch": 0.713930760311413, + "grad_norm": 0.7528647184371948, + "learning_rate": 1.9989285972581595e-06, + "loss": 0.16419677734375, + "step": 2155, + "tIoU": 0.5330680847167969, + "tRecall_50": 0.5625, + "tRecall_75": 0.4, + "tRecall_90": 0.21875, + "token_acc": 0.9387254901960784 + }, + { + "epoch": 0.7155872121914858, + "grad_norm": 0.47473061084747314, + "learning_rate": 1.9775172320015805e-06, + "loss": 0.1602294921875, + "step": 2160, + "tIoU": 0.5205780506134033, + "tRecall_50": 0.58125, + "tRecall_75": 0.3875, + "tRecall_90": 0.2, + "token_acc": 0.9394843962008141 + }, + { + "epoch": 0.7172436640715587, + "grad_norm": 0.9584677219390869, + "learning_rate": 1.956192854956397e-06, + "loss": 0.17667236328125, + "step": 2165, + "tIoU": 0.47919406890869143, + "tRecall_50": 0.50625, + "tRecall_75": 0.38125, + "tRecall_90": 0.21875, + "token_acc": 0.9352750809061489 + }, + { + "epoch": 0.7189001159516316, + "grad_norm": 0.5655807852745056, + "learning_rate": 1.9349560798463423e-06, + "loss": 0.145880126953125, + "step": 2170, + "tIoU": 0.5800382137298584, + "tRecall_50": 0.6125, + "tRecall_75": 0.45, + "tRecall_90": 0.28125, + "token_acc": 0.9449164154562893 + }, + { + "epoch": 0.7205565678317045, + "grad_norm": 0.4855189323425293, + "learning_rate": 1.9138075178739375e-06, + "loss": 0.15904541015625, + "step": 2175, + "tIoU": 0.5785966873168945, + "tRecall_50": 0.6375, + "tRecall_75": 0.4625, + "tRecall_90": 0.31875, + "token_acc": 0.9443683200876952 + }, + { + "epoch": 0.7222130197117774, + "grad_norm": 0.42149361968040466, + "learning_rate": 1.8927477777028934e-06, + "loss": 0.16480712890625, + "step": 2180, + "tIoU": 0.5444125652313232, + "tRecall_50": 0.56875, + "tRecall_75": 0.40625, + "tRecall_90": 0.2625, + "token_acc": 0.9375678610206297 + }, + { + "epoch": 0.7238694715918502, + "grad_norm": 0.8240121006965637, + "learning_rate": 1.8717774654405962e-06, + "loss": 0.151251220703125, + "step": 2185, + "tIoU": 0.5565950393676757, + "tRecall_50": 0.6, + "tRecall_75": 0.41875, + "tRecall_90": 0.2875, + "token_acc": 0.9450013546464373 + }, + { + "epoch": 0.7255259234719231, + "grad_norm": 0.41935020685195923, + "learning_rate": 1.8508971846206603e-06, + "loss": 0.1607177734375, + "step": 2190, + "tIoU": 0.5410372257232666, + "tRecall_50": 0.575, + "tRecall_75": 0.40625, + "tRecall_90": 0.30625, + "token_acc": 0.9429812652728754 + }, + { + "epoch": 0.727182375351996, + "grad_norm": 0.6102359294891357, + "learning_rate": 1.8301075361855674e-06, + "loss": 0.1631591796875, + "step": 2195, + "tIoU": 0.49234714508056643, + "tRecall_50": 0.525, + "tRecall_75": 0.3375, + "tRecall_90": 0.25, + "token_acc": 0.9369517543859649 + }, + { + "epoch": 0.7288388272320689, + "grad_norm": 0.599194347858429, + "learning_rate": 1.8094091184693569e-06, + "loss": 0.1391357421875, + "step": 2200, + "tIoU": 0.5903083324432373, + "tRecall_50": 0.61875, + "tRecall_75": 0.49375, + "tRecall_90": 0.30625, + "token_acc": 0.9462336244541485 + }, + { + "epoch": 0.7304952791121417, + "grad_norm": 0.6685839295387268, + "learning_rate": 1.7888025271804156e-06, + "loss": 0.1526611328125, + "step": 2205, + "tIoU": 0.5808693408966065, + "tRecall_50": 0.60625, + "tRecall_75": 0.4375, + "tRecall_90": 0.3375, + "token_acc": 0.944063613929257 + }, + { + "epoch": 0.7321517309922146, + "grad_norm": 0.5242391228675842, + "learning_rate": 1.7682883553843356e-06, + "loss": 0.16241455078125, + "step": 2210, + "tIoU": 0.5209659576416016, + "tRecall_50": 0.5625, + "tRecall_75": 0.39375, + "tRecall_90": 0.24375, + "token_acc": 0.9373126192423004 + }, + { + "epoch": 0.7338081828722876, + "grad_norm": 0.6233892440795898, + "learning_rate": 1.7478671934868302e-06, + "loss": 0.16380615234375, + "step": 2215, + "tIoU": 0.5173926830291748, + "tRecall_50": 0.55625, + "tRecall_75": 0.3875, + "tRecall_90": 0.225, + "token_acc": 0.9400492745688476 + }, + { + "epoch": 0.7354646347523605, + "grad_norm": 0.8942482471466064, + "learning_rate": 1.7275396292167623e-06, + "loss": 0.1623779296875, + "step": 2220, + "tIoU": 0.5592443943023682, + "tRecall_50": 0.6, + "tRecall_75": 0.4, + "tRecall_90": 0.28125, + "token_acc": 0.943504366812227 + }, + { + "epoch": 0.7371210866324334, + "grad_norm": 0.7321513891220093, + "learning_rate": 1.7073062476092145e-06, + "loss": 0.174151611328125, + "step": 2225, + "tIoU": 0.4566510200500488, + "tRecall_50": 0.4875, + "tRecall_75": 0.33125, + "tRecall_90": 0.14375, + "token_acc": 0.9328804347826087 + }, + { + "epoch": 0.7387775385125062, + "grad_norm": 0.6549537181854248, + "learning_rate": 1.6871676309886548e-06, + "loss": 0.1498046875, + "step": 2230, + "tIoU": 0.5559914588928223, + "tRecall_50": 0.59375, + "tRecall_75": 0.41875, + "tRecall_90": 0.25, + "token_acc": 0.9419496166484118 + }, + { + "epoch": 0.7404339903925791, + "grad_norm": 0.6222982406616211, + "learning_rate": 1.667124358952184e-06, + "loss": 0.16259765625, + "step": 2235, + "tIoU": 0.5121729850769043, + "tRecall_50": 0.50625, + "tRecall_75": 0.35625, + "tRecall_90": 0.2375, + "token_acc": 0.9351851851851852 + }, + { + "epoch": 0.742090442272652, + "grad_norm": 0.6524860858917236, + "learning_rate": 1.6471770083528444e-06, + "loss": 0.16424560546875, + "step": 2240, + "tIoU": 0.5048126697540283, + "tRecall_50": 0.54375, + "tRecall_75": 0.35625, + "tRecall_90": 0.20625, + "token_acc": 0.9371273712737127 + }, + { + "epoch": 0.7437468941527249, + "grad_norm": 0.6570249199867249, + "learning_rate": 1.6273261532830242e-06, + "loss": 0.165899658203125, + "step": 2245, + "tIoU": 0.5277689933776856, + "tRecall_50": 0.55625, + "tRecall_75": 0.3625, + "tRecall_90": 0.25, + "token_acc": 0.9389479422185881 + }, + { + "epoch": 0.7454033460327978, + "grad_norm": 0.5734337568283081, + "learning_rate": 1.607572365057934e-06, + "loss": 0.15498046875, + "step": 2250, + "tIoU": 0.5635070323944091, + "tRecall_50": 0.625, + "tRecall_75": 0.475, + "tRecall_90": 0.2875, + "token_acc": 0.9411444141689373 + }, + { + "epoch": 0.7470597979128706, + "grad_norm": 0.5429418683052063, + "learning_rate": 1.5879162121991614e-06, + "loss": 0.148138427734375, + "step": 2255, + "tIoU": 0.5375117301940918, + "tRecall_50": 0.575, + "tRecall_75": 0.40625, + "tRecall_90": 0.24375, + "token_acc": 0.9419143714207799 + }, + { + "epoch": 0.7487162497929435, + "grad_norm": 0.4823099374771118, + "learning_rate": 1.5683582604183124e-06, + "loss": 0.1737548828125, + "step": 2260, + "tIoU": 0.4863481044769287, + "tRecall_50": 0.51875, + "tRecall_75": 0.3375, + "tRecall_90": 0.19375, + "token_acc": 0.9344262295081968 + }, + { + "epoch": 0.7503727016730164, + "grad_norm": 0.576435387134552, + "learning_rate": 1.5488990726007297e-06, + "loss": 0.159820556640625, + "step": 2265, + "tIoU": 0.5188336849212647, + "tRecall_50": 0.5375, + "tRecall_75": 0.33125, + "tRecall_90": 0.25, + "token_acc": 0.9416575790621592 + }, + { + "epoch": 0.7520291535530893, + "grad_norm": 0.7238029837608337, + "learning_rate": 1.5295392087892884e-06, + "loss": 0.1625244140625, + "step": 2270, + "tIoU": 0.5586510181427002, + "tRecall_50": 0.575, + "tRecall_75": 0.36875, + "tRecall_90": 0.25625, + "token_acc": 0.9395008138903961 + }, + { + "epoch": 0.7536856054331622, + "grad_norm": 0.6263623833656311, + "learning_rate": 1.5102792261682813e-06, + "loss": 0.16297607421875, + "step": 2275, + "tIoU": 0.536182975769043, + "tRecall_50": 0.575, + "tRecall_75": 0.375, + "tRecall_90": 0.26875, + "token_acc": 0.9403107113654947 + }, + { + "epoch": 0.755342057313235, + "grad_norm": 0.6171078681945801, + "learning_rate": 1.4911196790473876e-06, + "loss": 0.17288818359375, + "step": 2280, + "tIoU": 0.510650634765625, + "tRecall_50": 0.525, + "tRecall_75": 0.3875, + "tRecall_90": 0.25625, + "token_acc": 0.9331340038053819 + }, + { + "epoch": 0.7569985091933079, + "grad_norm": 0.5526485443115234, + "learning_rate": 1.472061118845703e-06, + "loss": 0.15936279296875, + "step": 2285, + "tIoU": 0.5104878902435303, + "tRecall_50": 0.525, + "tRecall_75": 0.3875, + "tRecall_90": 0.23125, + "token_acc": 0.9381639880141651 + }, + { + "epoch": 0.7586549610733808, + "grad_norm": 0.7188597321510315, + "learning_rate": 1.4531040940758922e-06, + "loss": 0.14937744140625, + "step": 2290, + "tIoU": 0.5513885498046875, + "tRecall_50": 0.63125, + "tRecall_75": 0.4125, + "tRecall_90": 0.2625, + "token_acc": 0.9457553566585299 + }, + { + "epoch": 0.7603114129534537, + "grad_norm": 0.5185531973838806, + "learning_rate": 1.4342491503283862e-06, + "loss": 0.16483154296875, + "step": 2295, + "tIoU": 0.5377781391143799, + "tRecall_50": 0.5875, + "tRecall_75": 0.43125, + "tRecall_90": 0.275, + "token_acc": 0.9410474362489718 + }, + { + "epoch": 0.7619678648335266, + "grad_norm": 1.1246395111083984, + "learning_rate": 1.415496830255685e-06, + "loss": 0.156219482421875, + "step": 2300, + "tIoU": 0.5359954833984375, + "tRecall_50": 0.59375, + "tRecall_75": 0.3875, + "tRecall_90": 0.26875, + "token_acc": 0.9422185881711638 + }, + { + "epoch": 0.7636243167135994, + "grad_norm": 0.6118795275688171, + "learning_rate": 1.3968476735567392e-06, + "loss": 0.16826171875, + "step": 2305, + "tIoU": 0.5206178665161133, + "tRecall_50": 0.5625, + "tRecall_75": 0.3875, + "tRecall_90": 0.2625, + "token_acc": 0.9380965366784838 + }, + { + "epoch": 0.7652807685936723, + "grad_norm": 0.6857731342315674, + "learning_rate": 1.378302216961423e-06, + "loss": 0.1671875, + "step": 2310, + "tIoU": 0.5181430816650391, + "tRecall_50": 0.55625, + "tRecall_75": 0.35625, + "tRecall_90": 0.2125, + "token_acc": 0.9378055404671374 + }, + { + "epoch": 0.7669372204737452, + "grad_norm": 0.5392675995826721, + "learning_rate": 1.3598609942150765e-06, + "loss": 0.15540771484375, + "step": 2315, + "tIoU": 0.547668170928955, + "tRecall_50": 0.6125, + "tRecall_75": 0.3875, + "tRecall_90": 0.24375, + "token_acc": 0.9385245901639344 + }, + { + "epoch": 0.7685936723538181, + "grad_norm": 0.6362417936325073, + "learning_rate": 1.3415245360631519e-06, + "loss": 0.167828369140625, + "step": 2320, + "tIoU": 0.5249619960784913, + "tRecall_50": 0.56875, + "tRecall_75": 0.39375, + "tRecall_90": 0.225, + "token_acc": 0.9399782726778925 + }, + { + "epoch": 0.770250124233891, + "grad_norm": 0.7602875232696533, + "learning_rate": 1.3232933702359357e-06, + "loss": 0.1555908203125, + "step": 2325, + "tIoU": 0.5604390621185302, + "tRecall_50": 0.6125, + "tRecall_75": 0.40625, + "tRecall_90": 0.2875, + "token_acc": 0.9440959912735206 + }, + { + "epoch": 0.7719065761139638, + "grad_norm": 0.6373392939567566, + "learning_rate": 1.30516802143336e-06, + "loss": 0.15029296875, + "step": 2330, + "tIoU": 0.5182899951934814, + "tRecall_50": 0.55, + "tRecall_75": 0.38125, + "tRecall_90": 0.2375, + "token_acc": 0.9408235614944096 + }, + { + "epoch": 0.7735630279940368, + "grad_norm": 0.7477315068244934, + "learning_rate": 1.2871490113099066e-06, + "loss": 0.1555419921875, + "step": 2335, + "tIoU": 0.5171179294586181, + "tRecall_50": 0.53125, + "tRecall_75": 0.35, + "tRecall_90": 0.20625, + "token_acc": 0.9402660874287265 + }, + { + "epoch": 0.7752194798741097, + "grad_norm": 0.789494514465332, + "learning_rate": 1.269236858459585e-06, + "loss": 0.15079345703125, + "step": 2340, + "tIoU": 0.5262135982513427, + "tRecall_50": 0.56875, + "tRecall_75": 0.36875, + "tRecall_90": 0.2375, + "token_acc": 0.9411603721948549 + }, + { + "epoch": 0.7768759317541826, + "grad_norm": 0.5260175466537476, + "learning_rate": 1.251432078401011e-06, + "loss": 0.15489501953125, + "step": 2345, + "tIoU": 0.5641520500183106, + "tRecall_50": 0.61875, + "tRecall_75": 0.3875, + "tRecall_90": 0.25625, + "token_acc": 0.9441431670281996 + }, + { + "epoch": 0.7785323836342555, + "grad_norm": 0.6776566505432129, + "learning_rate": 1.2337351835625772e-06, + "loss": 0.16513671875, + "step": 2350, + "tIoU": 0.5146319389343261, + "tRecall_50": 0.575, + "tRecall_75": 0.4, + "tRecall_90": 0.2375, + "token_acc": 0.9373461012311901 + }, + { + "epoch": 0.7801888355143283, + "grad_norm": 0.7546380162239075, + "learning_rate": 1.2161466832676888e-06, + "loss": 0.16092529296875, + "step": 2355, + "tIoU": 0.5249964714050293, + "tRecall_50": 0.5625, + "tRecall_75": 0.40625, + "tRecall_90": 0.23125, + "token_acc": 0.941304347826087 + }, + { + "epoch": 0.7818452873944012, + "grad_norm": 0.5215368866920471, + "learning_rate": 1.1986670837201186e-06, + "loss": 0.14600830078125, + "step": 2360, + "tIoU": 0.5563305854797364, + "tRecall_50": 0.6, + "tRecall_75": 0.40625, + "tRecall_90": 0.3, + "token_acc": 0.9457617879531207 + }, + { + "epoch": 0.7835017392744741, + "grad_norm": 0.7974917888641357, + "learning_rate": 1.1812968879894387e-06, + "loss": 0.172119140625, + "step": 2365, + "tIoU": 0.5147394657135009, + "tRecall_50": 0.575, + "tRecall_75": 0.34375, + "tRecall_90": 0.1625, + "token_acc": 0.9330976339407125 + }, + { + "epoch": 0.785158191154547, + "grad_norm": 0.39690589904785156, + "learning_rate": 1.1640365959965305e-06, + "loss": 0.16275634765625, + "step": 2370, + "tIoU": 0.5097333431243897, + "tRecall_50": 0.58125, + "tRecall_75": 0.35, + "tRecall_90": 0.2125, + "token_acc": 0.9370420624151967 + }, + { + "epoch": 0.7868146430346199, + "grad_norm": 0.48728054761886597, + "learning_rate": 1.1468867044992065e-06, + "loss": 0.15360107421875, + "step": 2375, + "tIoU": 0.548112154006958, + "tRecall_50": 0.575, + "tRecall_75": 0.4, + "tRecall_90": 0.28125, + "token_acc": 0.9423024336888159 + }, + { + "epoch": 0.7884710949146927, + "grad_norm": 0.642907977104187, + "learning_rate": 1.1298477070779123e-06, + "loss": 0.153314208984375, + "step": 2380, + "tIoU": 0.5524973392486572, + "tRecall_50": 0.59375, + "tRecall_75": 0.4125, + "tRecall_90": 0.26875, + "token_acc": 0.9427408412483039 + }, + { + "epoch": 0.7901275467947656, + "grad_norm": 0.4547934830188751, + "learning_rate": 1.1129200941215185e-06, + "loss": 0.1518798828125, + "step": 2385, + "tIoU": 0.5372862339019775, + "tRecall_50": 0.5875, + "tRecall_75": 0.39375, + "tRecall_90": 0.2375, + "token_acc": 0.9394188596491229 + }, + { + "epoch": 0.7917839986748385, + "grad_norm": 0.45014676451683044, + "learning_rate": 1.096104352813206e-06, + "loss": 0.16217041015625, + "step": 2390, + "tIoU": 0.5642598152160645, + "tRecall_50": 0.63125, + "tRecall_75": 0.41875, + "tRecall_90": 0.3125, + "token_acc": 0.9434117003827228 + }, + { + "epoch": 0.7934404505549114, + "grad_norm": 0.584336519241333, + "learning_rate": 1.0794009671164484e-06, + "loss": 0.1718505859375, + "step": 2395, + "tIoU": 0.503229808807373, + "tRecall_50": 0.56875, + "tRecall_75": 0.3125, + "tRecall_90": 0.2, + "token_acc": 0.9349371241115364 + }, + { + "epoch": 0.7950969024349842, + "grad_norm": 0.7591623663902283, + "learning_rate": 1.062810417761082e-06, + "loss": 0.15487060546875, + "step": 2400, + "tIoU": 0.5614675521850586, + "tRecall_50": 0.6, + "tRecall_75": 0.41875, + "tRecall_90": 0.3, + "token_acc": 0.9457322061630761 + }, + { + "epoch": 0.7967533543150571, + "grad_norm": 0.6303367018699646, + "learning_rate": 1.046333182229472e-06, + "loss": 0.16737060546875, + "step": 2405, + "tIoU": 0.4994485855102539, + "tRecall_50": 0.55625, + "tRecall_75": 0.3375, + "tRecall_90": 0.21875, + "token_acc": 0.9360892031547458 + }, + { + "epoch": 0.79840980619513, + "grad_norm": 0.8385005593299866, + "learning_rate": 1.0299697347427645e-06, + "loss": 0.1697509765625, + "step": 2410, + "tIoU": 0.554076862335205, + "tRecall_50": 0.625, + "tRecall_75": 0.38125, + "tRecall_90": 0.25, + "token_acc": 0.9413689664575947 + }, + { + "epoch": 0.8000662580752029, + "grad_norm": 0.6249813437461853, + "learning_rate": 1.0137205462472427e-06, + "loss": 0.14515380859375, + "step": 2415, + "tIoU": 0.5932907581329345, + "tRecall_50": 0.66875, + "tRecall_75": 0.4625, + "tRecall_90": 0.28125, + "token_acc": 0.9468638729115311 + }, + { + "epoch": 0.8017227099552758, + "grad_norm": 2.346714973449707, + "learning_rate": 9.975860844007774e-07, + "loss": 0.16220703125, + "step": 2420, + "tIoU": 0.47458863258361816, + "tRecall_50": 0.51875, + "tRecall_75": 0.31875, + "tRecall_90": 0.20625, + "token_acc": 0.9378068739770867 + }, + { + "epoch": 0.8033791618353486, + "grad_norm": 0.6921612620353699, + "learning_rate": 9.815668135593548e-07, + "loss": 0.1552490234375, + "step": 2425, + "tIoU": 0.5383291244506836, + "tRecall_50": 0.54375, + "tRecall_75": 0.4, + "tRecall_90": 0.3, + "token_acc": 0.9442622950819672 + }, + { + "epoch": 0.8050356137154215, + "grad_norm": 0.7046741843223572, + "learning_rate": 9.656631947637235e-07, + "loss": 0.15458984375, + "step": 2430, + "tIoU": 0.5690701007843018, + "tRecall_50": 0.60625, + "tRecall_75": 0.45625, + "tRecall_90": 0.25, + "token_acc": 0.9448669201520913 + }, + { + "epoch": 0.8066920655954944, + "grad_norm": 0.5167131423950195, + "learning_rate": 9.498756857261243e-07, + "loss": 0.15469970703125, + "step": 2435, + "tIoU": 0.5229488849639893, + "tRecall_50": 0.5625, + "tRecall_75": 0.4, + "tRecall_90": 0.26875, + "token_acc": 0.9429273620972146 + }, + { + "epoch": 0.8083485174755674, + "grad_norm": 0.472789466381073, + "learning_rate": 9.342047408171117e-07, + "loss": 0.167047119140625, + "step": 2440, + "tIoU": 0.5211413860321045, + "tRecall_50": 0.56875, + "tRecall_75": 0.4, + "tRecall_90": 0.25, + "token_acc": 0.9389146441232615 + }, + { + "epoch": 0.8100049693556403, + "grad_norm": 0.4715271294116974, + "learning_rate": 9.186508110524795e-07, + "loss": 0.15396728515625, + "step": 2445, + "tIoU": 0.5372179031372071, + "tRecall_50": 0.56875, + "tRecall_75": 0.425, + "tRecall_90": 0.26875, + "token_acc": 0.942888522789676 + }, + { + "epoch": 0.811661421235713, + "grad_norm": 0.5346218347549438, + "learning_rate": 9.032143440802837e-07, + "loss": 0.155706787109375, + "step": 2450, + "tIoU": 0.5261980533599854, + "tRecall_50": 0.5625, + "tRecall_75": 0.40625, + "tRecall_90": 0.28125, + "token_acc": 0.941464742717125 + }, + { + "epoch": 0.813317873115786, + "grad_norm": 0.5118564367294312, + "learning_rate": 8.878957841679542e-07, + "loss": 0.15142822265625, + "step": 2455, + "tIoU": 0.5573920249938965, + "tRecall_50": 0.625, + "tRecall_75": 0.4125, + "tRecall_90": 0.2625, + "token_acc": 0.9446103719793646 + }, + { + "epoch": 0.8149743249958589, + "grad_norm": 0.5679037570953369, + "learning_rate": 8.726955721895103e-07, + "loss": 0.15771484375, + "step": 2460, + "tIoU": 0.5401461601257325, + "tRecall_50": 0.5875, + "tRecall_75": 0.425, + "tRecall_90": 0.2875, + "token_acc": 0.9441707717569786 + }, + { + "epoch": 0.8166307768759318, + "grad_norm": 0.5475548505783081, + "learning_rate": 8.576141456128734e-07, + "loss": 0.16822509765625, + "step": 2465, + "tIoU": 0.4817670345306396, + "tRecall_50": 0.5, + "tRecall_75": 0.35, + "tRecall_90": 0.225, + "token_acc": 0.9360830374214696 + }, + { + "epoch": 0.8182872287560047, + "grad_norm": 0.49165377020835876, + "learning_rate": 8.426519384872733e-07, + "loss": 0.1654541015625, + "step": 2470, + "tIoU": 0.5372585773468017, + "tRecall_50": 0.59375, + "tRecall_75": 0.41875, + "tRecall_90": 0.2625, + "token_acc": 0.9378762999452654 + }, + { + "epoch": 0.8199436806360775, + "grad_norm": 0.6567497253417969, + "learning_rate": 8.278093814307637e-07, + "loss": 0.1543701171875, + "step": 2475, + "tIoU": 0.5190737247467041, + "tRecall_50": 0.56875, + "tRecall_75": 0.4125, + "tRecall_90": 0.225, + "token_acc": 0.9400656814449918 + }, + { + "epoch": 0.8216001325161504, + "grad_norm": 0.6542906761169434, + "learning_rate": 8.130869016178178e-07, + "loss": 0.15341796875, + "step": 2480, + "tIoU": 0.5407592296600342, + "tRecall_50": 0.60625, + "tRecall_75": 0.38125, + "tRecall_90": 0.2625, + "token_acc": 0.9434889434889435 + }, + { + "epoch": 0.8232565843962233, + "grad_norm": 0.6917836666107178, + "learning_rate": 7.984849227670421e-07, + "loss": 0.162042236328125, + "step": 2485, + "tIoU": 0.5039559364318847, + "tRecall_50": 0.53125, + "tRecall_75": 0.3375, + "tRecall_90": 0.2375, + "token_acc": 0.9401455918037207 + }, + { + "epoch": 0.8249130362762962, + "grad_norm": 0.6990655660629272, + "learning_rate": 7.840038651289794e-07, + "loss": 0.17220458984375, + "step": 2490, + "tIoU": 0.5036819458007813, + "tRecall_50": 0.5375, + "tRecall_75": 0.35, + "tRecall_90": 0.25625, + "token_acc": 0.9378898833740168 + }, + { + "epoch": 0.8265694881563691, + "grad_norm": 0.6285939812660217, + "learning_rate": 7.696441454740134e-07, + "loss": 0.161029052734375, + "step": 2495, + "tIoU": 0.5125839710235596, + "tRecall_50": 0.55, + "tRecall_75": 0.33125, + "tRecall_90": 0.23125, + "token_acc": 0.9377559377559378 + }, + { + "epoch": 0.8282259400364419, + "grad_norm": 0.6815149188041687, + "learning_rate": 7.554061770803733e-07, + "loss": 0.16112060546875, + "step": 2500, + "tIoU": 0.5246678352355957, + "tRecall_50": 0.58125, + "tRecall_75": 0.375, + "tRecall_90": 0.21875, + "token_acc": 0.9360260233125508 + }, + { + "epoch": 0.8298823919165148, + "grad_norm": 0.7072858810424805, + "learning_rate": 7.412903697222429e-07, + "loss": 0.1623291015625, + "step": 2505, + "tIoU": 0.5076375961303711, + "tRecall_50": 0.5125, + "tRecall_75": 0.38125, + "tRecall_90": 0.225, + "token_acc": 0.9397163120567376 + }, + { + "epoch": 0.8315388437965877, + "grad_norm": 0.5564939379692078, + "learning_rate": 7.272971296579629e-07, + "loss": 0.15618896484375, + "step": 2510, + "tIoU": 0.5498382568359375, + "tRecall_50": 0.6, + "tRecall_75": 0.4125, + "tRecall_90": 0.21875, + "token_acc": 0.9418732782369146 + }, + { + "epoch": 0.8331952956766606, + "grad_norm": 0.5424681901931763, + "learning_rate": 7.13426859618338e-07, + "loss": 0.15855712890625, + "step": 2515, + "tIoU": 0.5336521625518799, + "tRecall_50": 0.58125, + "tRecall_75": 0.3875, + "tRecall_90": 0.225, + "token_acc": 0.9439277899343544 + }, + { + "epoch": 0.8348517475567335, + "grad_norm": 0.48574844002723694, + "learning_rate": 6.996799587950542e-07, + "loss": 0.165234375, + "step": 2520, + "tIoU": 0.5291247367858887, + "tRecall_50": 0.55, + "tRecall_75": 0.40625, + "tRecall_90": 0.2875, + "token_acc": 0.9401639344262295 + }, + { + "epoch": 0.8365081994368063, + "grad_norm": 0.656525194644928, + "learning_rate": 6.860568228291808e-07, + "loss": 0.153076171875, + "step": 2525, + "tIoU": 0.5800621509552002, + "tRecall_50": 0.6375, + "tRecall_75": 0.45, + "tRecall_90": 0.26875, + "token_acc": 0.9424440807419531 + }, + { + "epoch": 0.8381646513168792, + "grad_norm": 0.7389978766441345, + "learning_rate": 6.725578437997837e-07, + "loss": 0.15035400390625, + "step": 2530, + "tIoU": 0.5640317916870117, + "tRecall_50": 0.6125, + "tRecall_75": 0.425, + "tRecall_90": 0.2625, + "token_acc": 0.9400328587075575 + }, + { + "epoch": 0.8398211031969521, + "grad_norm": 0.7263797521591187, + "learning_rate": 6.591834102126516e-07, + "loss": 0.1676513671875, + "step": 2535, + "tIoU": 0.49884805679321287, + "tRecall_50": 0.53125, + "tRecall_75": 0.34375, + "tRecall_90": 0.2375, + "token_acc": 0.9385869565217392 + }, + { + "epoch": 0.841477555077025, + "grad_norm": 0.6837959885597229, + "learning_rate": 6.459339069891036e-07, + "loss": 0.1603759765625, + "step": 2540, + "tIoU": 0.5144101619720459, + "tRecall_50": 0.50625, + "tRecall_75": 0.3875, + "tRecall_90": 0.2875, + "token_acc": 0.9393442622950819 + }, + { + "epoch": 0.843134006957098, + "grad_norm": 0.5565042495727539, + "learning_rate": 6.328097154549146e-07, + "loss": 0.1583740234375, + "step": 2545, + "tIoU": 0.5358321666717529, + "tRecall_50": 0.56875, + "tRecall_75": 0.375, + "tRecall_90": 0.23125, + "token_acc": 0.9405860010851872 + }, + { + "epoch": 0.8447904588371707, + "grad_norm": 0.4988414943218231, + "learning_rate": 6.198112133293438e-07, + "loss": 0.17939453125, + "step": 2550, + "tIoU": 0.5300264835357666, + "tRecall_50": 0.5375, + "tRecall_75": 0.40625, + "tRecall_90": 0.28125, + "token_acc": 0.9378746594005449 + }, + { + "epoch": 0.8464469107172437, + "grad_norm": 0.6543470621109009, + "learning_rate": 6.069387747142591e-07, + "loss": 0.16768798828125, + "step": 2555, + "tIoU": 0.5053749084472656, + "tRecall_50": 0.51875, + "tRecall_75": 0.3625, + "tRecall_90": 0.28125, + "token_acc": 0.93435031326614 + }, + { + "epoch": 0.8481033625973166, + "grad_norm": 0.9791786074638367, + "learning_rate": 5.941927700833727e-07, + "loss": 0.16727294921875, + "step": 2560, + "tIoU": 0.5062736988067627, + "tRecall_50": 0.55625, + "tRecall_75": 0.29375, + "tRecall_90": 0.2125, + "token_acc": 0.937363238512035 + }, + { + "epoch": 0.8497598144773895, + "grad_norm": 0.6862460374832153, + "learning_rate": 5.815735662715782e-07, + "loss": 0.164141845703125, + "step": 2565, + "tIoU": 0.5086766719818115, + "tRecall_50": 0.54375, + "tRecall_75": 0.375, + "tRecall_90": 0.25625, + "token_acc": 0.9397491821155943 + }, + { + "epoch": 0.8514162663574624, + "grad_norm": 0.5261525511741638, + "learning_rate": 5.690815264643929e-07, + "loss": 0.14600830078125, + "step": 2570, + "tIoU": 0.5660731315612793, + "tRecall_50": 0.6, + "tRecall_75": 0.41875, + "tRecall_90": 0.2625, + "token_acc": 0.9455189321710705 + }, + { + "epoch": 0.8530727182375352, + "grad_norm": 0.6874206066131592, + "learning_rate": 5.567170101875074e-07, + "loss": 0.16983642578125, + "step": 2575, + "tIoU": 0.5341813087463378, + "tRecall_50": 0.5625, + "tRecall_75": 0.40625, + "tRecall_90": 0.2875, + "token_acc": 0.9367396593673966 + }, + { + "epoch": 0.8547291701176081, + "grad_norm": 0.5525768399238586, + "learning_rate": 5.444803732964333e-07, + "loss": 0.16873779296875, + "step": 2580, + "tIoU": 0.4944866180419922, + "tRecall_50": 0.5625, + "tRecall_75": 0.33125, + "tRecall_90": 0.16875, + "token_acc": 0.937244201909959 + }, + { + "epoch": 0.856385621997681, + "grad_norm": 0.5512920618057251, + "learning_rate": 5.323719679662647e-07, + "loss": 0.172900390625, + "step": 2585, + "tIoU": 0.5066828250885009, + "tRecall_50": 0.55625, + "tRecall_75": 0.35625, + "tRecall_90": 0.225, + "token_acc": 0.9372784292337061 + }, + { + "epoch": 0.8580420738777539, + "grad_norm": 1.3920177221298218, + "learning_rate": 5.203921426815467e-07, + "loss": 0.174609375, + "step": 2590, + "tIoU": 0.528574800491333, + "tRecall_50": 0.56875, + "tRecall_75": 0.3875, + "tRecall_90": 0.26875, + "token_acc": 0.9373803664205633 + }, + { + "epoch": 0.8596985257578267, + "grad_norm": 0.6287409663200378, + "learning_rate": 5.085412422262364e-07, + "loss": 0.17049560546875, + "step": 2595, + "tIoU": 0.4851510047912598, + "tRecall_50": 0.51875, + "tRecall_75": 0.33125, + "tRecall_90": 0.1625, + "token_acc": 0.931588988825293 + }, + { + "epoch": 0.8613549776378996, + "grad_norm": 0.5815898776054382, + "learning_rate": 4.968196076737847e-07, + "loss": 0.152935791015625, + "step": 2600, + "tIoU": 0.5580605030059814, + "tRecall_50": 0.625, + "tRecall_75": 0.39375, + "tRecall_90": 0.24375, + "token_acc": 0.9416278432447246 + }, + { + "epoch": 0.8630114295179725, + "grad_norm": 0.667481541633606, + "learning_rate": 4.852275763773251e-07, + "loss": 0.15257568359375, + "step": 2605, + "tIoU": 0.5585962295532226, + "tRecall_50": 0.61875, + "tRecall_75": 0.375, + "tRecall_90": 0.2375, + "token_acc": 0.9419302071973827 + }, + { + "epoch": 0.8646678813980454, + "grad_norm": 0.5576724410057068, + "learning_rate": 4.737654819599563e-07, + "loss": 0.1552734375, + "step": 2610, + "tIoU": 0.5506963729858398, + "tRecall_50": 0.60625, + "tRecall_75": 0.4125, + "tRecall_90": 0.28125, + "token_acc": 0.9373798407031035 + }, + { + "epoch": 0.8663243332781183, + "grad_norm": 0.45226308703422546, + "learning_rate": 4.624336543051433e-07, + "loss": 0.1733154296875, + "step": 2615, + "tIoU": 0.4877234935760498, + "tRecall_50": 0.53125, + "tRecall_75": 0.35, + "tRecall_90": 0.175, + "token_acc": 0.9375338020551649 + }, + { + "epoch": 0.8679807851581911, + "grad_norm": 0.757668137550354, + "learning_rate": 4.5123241954722673e-07, + "loss": 0.154150390625, + "step": 2620, + "tIoU": 0.5472731113433837, + "tRecall_50": 0.575, + "tRecall_75": 0.41875, + "tRecall_90": 0.28125, + "token_acc": 0.9441128594682583 + }, + { + "epoch": 0.869637237038264, + "grad_norm": 0.6104984879493713, + "learning_rate": 4.4016210006203096e-07, + "loss": 0.1505615234375, + "step": 2625, + "tIoU": 0.5675601005554199, + "tRecall_50": 0.56875, + "tRecall_75": 0.45625, + "tRecall_90": 0.30625, + "token_acc": 0.9440959912735206 + }, + { + "epoch": 0.8712936889183369, + "grad_norm": 0.6254884004592896, + "learning_rate": 4.292230144575904e-07, + "loss": 0.15528564453125, + "step": 2630, + "tIoU": 0.5098196029663086, + "tRecall_50": 0.5375, + "tRecall_75": 0.40625, + "tRecall_90": 0.225, + "token_acc": 0.9385034013605442 + }, + { + "epoch": 0.8729501407984098, + "grad_norm": 0.8128376603126526, + "learning_rate": 4.184154775649768e-07, + "loss": 0.164764404296875, + "step": 2635, + "tIoU": 0.5138954162597656, + "tRecall_50": 0.5625, + "tRecall_75": 0.4, + "tRecall_90": 0.20625, + "token_acc": 0.9365986394557824 + }, + { + "epoch": 0.8746065926784827, + "grad_norm": 0.5039224028587341, + "learning_rate": 4.0773980042923986e-07, + "loss": 0.148931884765625, + "step": 2640, + "tIoU": 0.5136477947235107, + "tRecall_50": 0.53125, + "tRecall_75": 0.3625, + "tRecall_90": 0.25, + "token_acc": 0.9414951245937161 + }, + { + "epoch": 0.8762630445585555, + "grad_norm": 0.4894895553588867, + "learning_rate": 3.9719629030045616e-07, + "loss": 0.17333984375, + "step": 2645, + "tIoU": 0.469511604309082, + "tRecall_50": 0.4625, + "tRecall_75": 0.3125, + "tRecall_90": 0.18125, + "token_acc": 0.9365468409586056 + }, + { + "epoch": 0.8779194964386284, + "grad_norm": 0.4093848168849945, + "learning_rate": 3.867852506248843e-07, + "loss": 0.16063232421875, + "step": 2650, + "tIoU": 0.49820685386657715, + "tRecall_50": 0.5375, + "tRecall_75": 0.35625, + "tRecall_90": 0.24375, + "token_acc": 0.9370552818828681 + }, + { + "epoch": 0.8795759483187013, + "grad_norm": 0.5204543471336365, + "learning_rate": 3.765069810362326e-07, + "loss": 0.166009521484375, + "step": 2655, + "tIoU": 0.5327613353729248, + "tRecall_50": 0.5625, + "tRecall_75": 0.3875, + "tRecall_90": 0.24375, + "token_acc": 0.9399890889252591 + }, + { + "epoch": 0.8812324001987742, + "grad_norm": 0.6944509148597717, + "learning_rate": 3.6636177734703405e-07, + "loss": 0.15164794921875, + "step": 2660, + "tIoU": 0.5492005825042725, + "tRecall_50": 0.60625, + "tRecall_75": 0.40625, + "tRecall_90": 0.25, + "token_acc": 0.9418003807451727 + }, + { + "epoch": 0.8828888520788472, + "grad_norm": 0.7271901965141296, + "learning_rate": 3.563499315401386e-07, + "loss": 0.145611572265625, + "step": 2665, + "tIoU": 0.5534201622009277, + "tRecall_50": 0.6125, + "tRecall_75": 0.425, + "tRecall_90": 0.30625, + "token_acc": 0.9439608269858542 + }, + { + "epoch": 0.88454530395892, + "grad_norm": 0.6131239533424377, + "learning_rate": 3.4647173176029835e-07, + "loss": 0.1537109375, + "step": 2670, + "tIoU": 0.5357179164886474, + "tRecall_50": 0.575, + "tRecall_75": 0.4125, + "tRecall_90": 0.26875, + "token_acc": 0.9426162632580908 + }, + { + "epoch": 0.8862017558389929, + "grad_norm": 0.49681997299194336, + "learning_rate": 3.3672746230588803e-07, + "loss": 0.15107421875, + "step": 2675, + "tIoU": 0.576555585861206, + "tRecall_50": 0.63125, + "tRecall_75": 0.43125, + "tRecall_90": 0.25625, + "token_acc": 0.9429192715411797 + }, + { + "epoch": 0.8878582077190658, + "grad_norm": 0.5076614618301392, + "learning_rate": 3.271174036207131e-07, + "loss": 0.158447265625, + "step": 2680, + "tIoU": 0.5341426849365234, + "tRecall_50": 0.55, + "tRecall_75": 0.375, + "tRecall_90": 0.25625, + "token_acc": 0.9407125373946151 + }, + { + "epoch": 0.8895146595991387, + "grad_norm": 0.6126746535301208, + "learning_rate": 3.176418322859415e-07, + "loss": 0.1483642578125, + "step": 2685, + "tIoU": 0.5753330230712891, + "tRecall_50": 0.65, + "tRecall_75": 0.45625, + "tRecall_90": 0.3, + "token_acc": 0.941240776168352 + }, + { + "epoch": 0.8911711114792116, + "grad_norm": 0.7555765509605408, + "learning_rate": 3.0830102101214563e-07, + "loss": 0.177294921875, + "step": 2690, + "tIoU": 0.46550650596618653, + "tRecall_50": 0.48125, + "tRecall_75": 0.34375, + "tRecall_90": 0.2125, + "token_acc": 0.9303780255643187 + }, + { + "epoch": 0.8928275633592844, + "grad_norm": 0.7692764401435852, + "learning_rate": 2.990952386314505e-07, + "loss": 0.143023681640625, + "step": 2695, + "tIoU": 0.5630221366882324, + "tRecall_50": 0.625, + "tRecall_75": 0.425, + "tRecall_90": 0.3, + "token_acc": 0.944505194095134 + }, + { + "epoch": 0.8944840152393573, + "grad_norm": 0.5720428228378296, + "learning_rate": 2.900247500897974e-07, + "loss": 0.146484375, + "step": 2700, + "tIoU": 0.5490079879760742, + "tRecall_50": 0.6, + "tRecall_75": 0.4125, + "tRecall_90": 0.275, + "token_acc": 0.9426386233269598 + }, + { + "epoch": 0.8961404671194302, + "grad_norm": 0.8671318292617798, + "learning_rate": 2.8108981643932e-07, + "loss": 0.152166748046875, + "step": 2705, + "tIoU": 0.571419620513916, + "tRecall_50": 0.63125, + "tRecall_75": 0.41875, + "tRecall_90": 0.23125, + "token_acc": 0.9429041998353006 + }, + { + "epoch": 0.8977969189995031, + "grad_norm": 0.6312313079833984, + "learning_rate": 2.7229069483083015e-07, + "loss": 0.147705078125, + "step": 2710, + "tIoU": 0.569254732131958, + "tRecall_50": 0.61875, + "tRecall_75": 0.43125, + "tRecall_90": 0.26875, + "token_acc": 0.9452316076294278 + }, + { + "epoch": 0.899453370879576, + "grad_norm": 0.9674606919288635, + "learning_rate": 2.636276385064157e-07, + "loss": 0.15697021484375, + "step": 2715, + "tIoU": 0.5186336040496826, + "tRecall_50": 0.575, + "tRecall_75": 0.375, + "tRecall_90": 0.23125, + "token_acc": 0.9401360544217687 + }, + { + "epoch": 0.9011098227596488, + "grad_norm": 0.5326899290084839, + "learning_rate": 2.5510089679215646e-07, + "loss": 0.15101318359375, + "step": 2720, + "tIoU": 0.560830020904541, + "tRecall_50": 0.5875, + "tRecall_75": 0.45625, + "tRecall_90": 0.3, + "token_acc": 0.9469074986316366 + }, + { + "epoch": 0.9027662746397217, + "grad_norm": 0.510621964931488, + "learning_rate": 2.4671071509094367e-07, + "loss": 0.16943359375, + "step": 2725, + "tIoU": 0.489136266708374, + "tRecall_50": 0.53125, + "tRecall_75": 0.36875, + "tRecall_90": 0.19375, + "token_acc": 0.9332066250339397 + }, + { + "epoch": 0.9044227265197946, + "grad_norm": 0.6251363158226013, + "learning_rate": 2.3845733487541824e-07, + "loss": 0.169390869140625, + "step": 2730, + "tIoU": 0.4824803829193115, + "tRecall_50": 0.53125, + "tRecall_75": 0.31875, + "tRecall_90": 0.19375, + "token_acc": 0.9344083082809511 + }, + { + "epoch": 0.9060791783998675, + "grad_norm": 0.6128382086753845, + "learning_rate": 2.3034099368102525e-07, + "loss": 0.1482421875, + "step": 2735, + "tIoU": 0.5446168899536132, + "tRecall_50": 0.575, + "tRecall_75": 0.41875, + "tRecall_90": 0.2875, + "token_acc": 0.9440825190010858 + }, + { + "epoch": 0.9077356302799404, + "grad_norm": 1.0730916261672974, + "learning_rate": 2.2236192509916933e-07, + "loss": 0.159869384765625, + "step": 2740, + "tIoU": 0.4948151111602783, + "tRecall_50": 0.5375, + "tRecall_75": 0.3375, + "tRecall_90": 0.25, + "token_acc": 0.9408893709327549 + }, + { + "epoch": 0.9093920821600132, + "grad_norm": 0.5609004497528076, + "learning_rate": 2.1452035877050137e-07, + "loss": 0.15233154296875, + "step": 2745, + "tIoU": 0.5627793312072754, + "tRecall_50": 0.60625, + "tRecall_75": 0.425, + "tRecall_90": 0.325, + "token_acc": 0.9446413962367057 + }, + { + "epoch": 0.9110485340400861, + "grad_norm": 0.6590545773506165, + "learning_rate": 2.068165203783029e-07, + "loss": 0.153057861328125, + "step": 2750, + "tIoU": 0.5494290351867676, + "tRecall_50": 0.63125, + "tRecall_75": 0.35625, + "tRecall_90": 0.275, + "token_acc": 0.9417052574230454 + }, + { + "epoch": 0.912704985920159, + "grad_norm": 0.4576837122440338, + "learning_rate": 1.992506316419912e-07, + "loss": 0.1706298828125, + "step": 2755, + "tIoU": 0.5233198642730713, + "tRecall_50": 0.55, + "tRecall_75": 0.39375, + "tRecall_90": 0.23125, + "token_acc": 0.934640522875817 + }, + { + "epoch": 0.9143614378002319, + "grad_norm": 0.6967729330062866, + "learning_rate": 1.9182291031074275e-07, + "loss": 0.1559814453125, + "step": 2760, + "tIoU": 0.5519247055053711, + "tRecall_50": 0.6, + "tRecall_75": 0.4125, + "tRecall_90": 0.24375, + "token_acc": 0.9383783783783783 + }, + { + "epoch": 0.9160178896803048, + "grad_norm": 0.4412071406841278, + "learning_rate": 1.8453357015722097e-07, + "loss": 0.17286376953125, + "step": 2765, + "tIoU": 0.47763400077819823, + "tRecall_50": 0.5125, + "tRecall_75": 0.34375, + "tRecall_90": 0.225, + "token_acc": 0.9332971800433839 + }, + { + "epoch": 0.9176743415603776, + "grad_norm": 0.4622722864151001, + "learning_rate": 1.7738282097142622e-07, + "loss": 0.15433349609375, + "step": 2770, + "tIoU": 0.5353363513946533, + "tRecall_50": 0.575, + "tRecall_75": 0.4, + "tRecall_90": 0.24375, + "token_acc": 0.9383319750067917 + }, + { + "epoch": 0.9193307934404505, + "grad_norm": 0.7105121612548828, + "learning_rate": 1.7037086855465902e-07, + "loss": 0.1578369140625, + "step": 2775, + "tIoU": 0.5579864025115967, + "tRecall_50": 0.625, + "tRecall_75": 0.425, + "tRecall_90": 0.275, + "token_acc": 0.9409513395297977 + }, + { + "epoch": 0.9209872453205235, + "grad_norm": 0.516262948513031, + "learning_rate": 1.63497914713594e-07, + "loss": 0.18677978515625, + "step": 2780, + "tIoU": 0.4627009391784668, + "tRecall_50": 0.48125, + "tRecall_75": 0.35625, + "tRecall_90": 0.2375, + "token_acc": 0.93 + }, + { + "epoch": 0.9226436972005964, + "grad_norm": 0.5634688138961792, + "learning_rate": 1.5676415725447424e-07, + "loss": 0.15511474609375, + "step": 2785, + "tIoU": 0.5456976890563965, + "tRecall_50": 0.6, + "tRecall_75": 0.44375, + "tRecall_90": 0.2625, + "token_acc": 0.9393118396098619 + }, + { + "epoch": 0.9243001490806692, + "grad_norm": 0.648868978023529, + "learning_rate": 1.5016978997741903e-07, + "loss": 0.160699462890625, + "step": 2790, + "tIoU": 0.5549224853515625, + "tRecall_50": 0.6, + "tRecall_75": 0.41875, + "tRecall_90": 0.2625, + "token_acc": 0.9421239421239421 + }, + { + "epoch": 0.9259566009607421, + "grad_norm": 0.5581037998199463, + "learning_rate": 1.4371500267084337e-07, + "loss": 0.1560546875, + "step": 2795, + "tIoU": 0.5302978038787842, + "tRecall_50": 0.5875, + "tRecall_75": 0.3875, + "tRecall_90": 0.25, + "token_acc": 0.9379781420765028 + }, + { + "epoch": 0.927613052840815, + "grad_norm": 1.205271601676941, + "learning_rate": 1.3739998110599618e-07, + "loss": 0.16649169921875, + "step": 2800, + "tIoU": 0.5526116847991943, + "tRecall_50": 0.60625, + "tRecall_75": 0.4, + "tRecall_90": 0.24375, + "token_acc": 0.9417186229330442 + }, + { + "epoch": 0.9292695047208879, + "grad_norm": 0.5804457068443298, + "learning_rate": 1.3122490703161793e-07, + "loss": 0.14595947265625, + "step": 2805, + "tIoU": 0.5765332221984864, + "tRecall_50": 0.61875, + "tRecall_75": 0.4375, + "tRecall_90": 0.325, + "token_acc": 0.9430982847808331 + }, + { + "epoch": 0.9309259566009608, + "grad_norm": 0.5786797404289246, + "learning_rate": 1.2518995816870327e-07, + "loss": 0.16298828125, + "step": 2810, + "tIoU": 0.5211046218872071, + "tRecall_50": 0.5875, + "tRecall_75": 0.35, + "tRecall_90": 0.23125, + "token_acc": 0.9343995662781241 + }, + { + "epoch": 0.9325824084810336, + "grad_norm": 0.731719970703125, + "learning_rate": 1.192953082053927e-07, + "loss": 0.151806640625, + "step": 2815, + "tIoU": 0.5116126060485839, + "tRecall_50": 0.54375, + "tRecall_75": 0.38125, + "tRecall_90": 0.28125, + "token_acc": 0.9389208436044919 + }, + { + "epoch": 0.9342388603611065, + "grad_norm": 0.5219873785972595, + "learning_rate": 1.1354112679196894e-07, + "loss": 0.1593505859375, + "step": 2820, + "tIoU": 0.5594928741455079, + "tRecall_50": 0.5875, + "tRecall_75": 0.41875, + "tRecall_90": 0.3125, + "token_acc": 0.9441280173582859 + }, + { + "epoch": 0.9358953122411794, + "grad_norm": 0.6567935347557068, + "learning_rate": 1.0792757953597522e-07, + "loss": 0.161572265625, + "step": 2825, + "tIoU": 0.5246160507202149, + "tRecall_50": 0.6, + "tRecall_75": 0.3875, + "tRecall_90": 0.225, + "token_acc": 0.9388706140350878 + }, + { + "epoch": 0.9375517641212523, + "grad_norm": 0.6812028288841248, + "learning_rate": 1.0245482799745243e-07, + "loss": 0.15240478515625, + "step": 2830, + "tIoU": 0.5628941535949707, + "tRecall_50": 0.58125, + "tRecall_75": 0.425, + "tRecall_90": 0.29375, + "token_acc": 0.9423969423969424 + }, + { + "epoch": 0.9392082160013252, + "grad_norm": 0.49755436182022095, + "learning_rate": 9.712302968428567e-08, + "loss": 0.16417236328125, + "step": 2835, + "tIoU": 0.5263060092926025, + "tRecall_50": 0.58125, + "tRecall_75": 0.375, + "tRecall_90": 0.23125, + "token_acc": 0.9415300546448088 + }, + { + "epoch": 0.940864667881398, + "grad_norm": 0.6163895130157471, + "learning_rate": 9.193233804767054e-08, + "loss": 0.156787109375, + "step": 2840, + "tIoU": 0.5490593433380127, + "tRecall_50": 0.5875, + "tRecall_75": 0.4125, + "tRecall_90": 0.2875, + "token_acc": 0.9403757146746529 + }, + { + "epoch": 0.9425211197614709, + "grad_norm": 0.5530081987380981, + "learning_rate": 8.688290247770071e-08, + "loss": 0.1619384765625, + "step": 2845, + "tIoU": 0.5022723197937011, + "tRecall_50": 0.5375, + "tRecall_75": 0.36875, + "tRecall_90": 0.18125, + "token_acc": 0.9364303178484108 + }, + { + "epoch": 0.9441775716415438, + "grad_norm": 0.672563374042511, + "learning_rate": 8.197486829906564e-08, + "loss": 0.1586669921875, + "step": 2850, + "tIoU": 0.5501746654510498, + "tRecall_50": 0.61875, + "tRecall_75": 0.41875, + "tRecall_90": 0.18125, + "token_acc": 0.9395589436427988 + }, + { + "epoch": 0.9458340235216167, + "grad_norm": 0.45273101329803467, + "learning_rate": 7.720837676686743e-08, + "loss": 0.14420166015625, + "step": 2855, + "tIoU": 0.535654354095459, + "tRecall_50": 0.54375, + "tRecall_75": 0.44375, + "tRecall_90": 0.275, + "token_acc": 0.9438415626695605 + }, + { + "epoch": 0.9474904754016896, + "grad_norm": 0.5784236192703247, + "learning_rate": 7.258356506255781e-08, + "loss": 0.1594482421875, + "step": 2860, + "tIoU": 0.5448553085327148, + "tRecall_50": 0.58125, + "tRecall_75": 0.4625, + "tRecall_90": 0.3, + "token_acc": 0.9401639344262295 + }, + { + "epoch": 0.9491469272817624, + "grad_norm": 0.6065331697463989, + "learning_rate": 6.810056628998918e-08, + "loss": 0.15972900390625, + "step": 2865, + "tIoU": 0.49558448791503906, + "tRecall_50": 0.5, + "tRecall_75": 0.36875, + "tRecall_90": 0.21875, + "token_acc": 0.9342069342069342 + }, + { + "epoch": 0.9508033791618353, + "grad_norm": 0.547734797000885, + "learning_rate": 6.375950947158205e-08, + "loss": 0.15238037109375, + "step": 2870, + "tIoU": 0.5578720092773437, + "tRecall_50": 0.59375, + "tRecall_75": 0.425, + "tRecall_90": 0.25625, + "token_acc": 0.9423182066703116 + }, + { + "epoch": 0.9524598310419082, + "grad_norm": 0.5336986780166626, + "learning_rate": 5.9560519544614725e-08, + "loss": 0.15859375, + "step": 2875, + "tIoU": 0.5313952445983887, + "tRecall_50": 0.575, + "tRecall_75": 0.38125, + "tRecall_90": 0.275, + "token_acc": 0.9424283765347885 + }, + { + "epoch": 0.9541162829219811, + "grad_norm": 0.5611312985420227, + "learning_rate": 5.5503717357623945e-08, + "loss": 0.148968505859375, + "step": 2880, + "tIoU": 0.5653798103332519, + "tRecall_50": 0.61875, + "tRecall_75": 0.46875, + "tRecall_90": 0.3, + "token_acc": 0.9454892341237394 + }, + { + "epoch": 0.955772734802054, + "grad_norm": 0.33091646432876587, + "learning_rate": 5.158921966693042e-08, + "loss": 0.144146728515625, + "step": 2885, + "tIoU": 0.6037317752838135, + "tRecall_50": 0.675, + "tRecall_75": 0.4625, + "tRecall_90": 0.29375, + "token_acc": 0.9470362239297475 + }, + { + "epoch": 0.9574291866821268, + "grad_norm": 0.638655960559845, + "learning_rate": 4.7817139133277126e-08, + "loss": 0.1508544921875, + "step": 2890, + "tIoU": 0.5142153739929199, + "tRecall_50": 0.60625, + "tRecall_75": 0.3375, + "tRecall_90": 0.2, + "token_acc": 0.937192790824686 + }, + { + "epoch": 0.9590856385621997, + "grad_norm": 0.7477061152458191, + "learning_rate": 4.418758431858572e-08, + "loss": 0.160791015625, + "step": 2895, + "tIoU": 0.5251534938812256, + "tRecall_50": 0.575, + "tRecall_75": 0.3625, + "tRecall_90": 0.21875, + "token_acc": 0.9371584699453552 + }, + { + "epoch": 0.9607420904422727, + "grad_norm": 0.42086467146873474, + "learning_rate": 4.070065968283354e-08, + "loss": 0.15504150390625, + "step": 2900, + "tIoU": 0.571649169921875, + "tRecall_50": 0.625, + "tRecall_75": 0.43125, + "tRecall_90": 0.29375, + "token_acc": 0.9434271658923203 + }, + { + "epoch": 0.9623985423223456, + "grad_norm": 0.7915459871292114, + "learning_rate": 3.7356465581047105e-08, + "loss": 0.150555419921875, + "step": 2905, + "tIoU": 0.5531152248382568, + "tRecall_50": 0.6, + "tRecall_75": 0.40625, + "tRecall_90": 0.2875, + "token_acc": 0.9445203607543045 + }, + { + "epoch": 0.9640549942024185, + "grad_norm": 0.610054075717926, + "learning_rate": 3.4155098260412744e-08, + "loss": 0.158453369140625, + "step": 2910, + "tIoU": 0.5769317626953125, + "tRecall_50": 0.6375, + "tRecall_75": 0.45, + "tRecall_90": 0.28125, + "token_acc": 0.9402455661664393 + }, + { + "epoch": 0.9657114460824913, + "grad_norm": 0.5148957967758179, + "learning_rate": 3.1096649857508264e-08, + "loss": 0.16090087890625, + "step": 2915, + "tIoU": 0.5256228923797608, + "tRecall_50": 0.6, + "tRecall_75": 0.39375, + "tRecall_90": 0.225, + "token_acc": 0.9383821932681867 + }, + { + "epoch": 0.9673678979625642, + "grad_norm": 0.4604478180408478, + "learning_rate": 2.8181208395648973e-08, + "loss": 0.15223388671875, + "step": 2920, + "tIoU": 0.5296051502227783, + "tRecall_50": 0.58125, + "tRecall_75": 0.39375, + "tRecall_90": 0.23125, + "token_acc": 0.9399782726778925 + }, + { + "epoch": 0.9690243498426371, + "grad_norm": 0.5151129364967346, + "learning_rate": 2.5408857782357465e-08, + "loss": 0.173046875, + "step": 2925, + "tIoU": 0.47686147689819336, + "tRecall_50": 0.5, + "tRecall_75": 0.3375, + "tRecall_90": 0.2125, + "token_acc": 0.9381668946648427 + }, + { + "epoch": 0.97068080172271, + "grad_norm": 0.6586578488349915, + "learning_rate": 2.2779677806945567e-08, + "loss": 0.1605712890625, + "step": 2930, + "tIoU": 0.5430468559265137, + "tRecall_50": 0.59375, + "tRecall_75": 0.41875, + "tRecall_90": 0.25625, + "token_acc": 0.9390710382513662 + }, + { + "epoch": 0.9723372536027829, + "grad_norm": 0.752968966960907, + "learning_rate": 2.0293744138219495e-08, + "loss": 0.159130859375, + "step": 2935, + "tIoU": 0.5184082508087158, + "tRecall_50": 0.55625, + "tRecall_75": 0.36875, + "tRecall_90": 0.2125, + "token_acc": 0.9366984993178717 + }, + { + "epoch": 0.9739937054828557, + "grad_norm": 0.44599080085754395, + "learning_rate": 1.7951128322302168e-08, + "loss": 0.14085693359375, + "step": 2940, + "tIoU": 0.5751462936401367, + "tRecall_50": 0.625, + "tRecall_75": 0.4625, + "tRecall_90": 0.33125, + "token_acc": 0.9483747609942639 + }, + { + "epoch": 0.9756501573629286, + "grad_norm": 0.6227806210517883, + "learning_rate": 1.5751897780573734e-08, + "loss": 0.161328125, + "step": 2945, + "tIoU": 0.5526154994964599, + "tRecall_50": 0.60625, + "tRecall_75": 0.40625, + "tRecall_90": 0.28125, + "token_acc": 0.9422343324250682 + }, + { + "epoch": 0.9773066092430015, + "grad_norm": 0.5416595339775085, + "learning_rate": 1.3696115807732002e-08, + "loss": 0.15426025390625, + "step": 2950, + "tIoU": 0.5484757900238038, + "tRecall_50": 0.625, + "tRecall_75": 0.43125, + "tRecall_90": 0.24375, + "token_acc": 0.940677966101695 + }, + { + "epoch": 0.9789630611230744, + "grad_norm": 1.0069711208343506, + "learning_rate": 1.1783841569968368e-08, + "loss": 0.1627685546875, + "step": 2955, + "tIoU": 0.5158636093139648, + "tRecall_50": 0.575, + "tRecall_75": 0.35625, + "tRecall_90": 0.25625, + "token_acc": 0.9338255400601586 + }, + { + "epoch": 0.9806195130031473, + "grad_norm": 0.7520166039466858, + "learning_rate": 1.0015130103268044e-08, + "loss": 0.15572509765625, + "step": 2960, + "tIoU": 0.5539876461029053, + "tRecall_50": 0.58125, + "tRecall_75": 0.4, + "tRecall_90": 0.29375, + "token_acc": 0.9455391351943077 + }, + { + "epoch": 0.9822759648832201, + "grad_norm": 0.47164127230644226, + "learning_rate": 8.390032311824115e-09, + "loss": 0.1467529296875, + "step": 2965, + "tIoU": 0.5772971153259278, + "tRecall_50": 0.625, + "tRecall_75": 0.4125, + "tRecall_90": 0.30625, + "token_acc": 0.9446564885496184 + }, + { + "epoch": 0.983932416763293, + "grad_norm": 0.5869075655937195, + "learning_rate": 6.908594966573701e-09, + "loss": 0.16009521484375, + "step": 2970, + "tIoU": 0.5448875427246094, + "tRecall_50": 0.60625, + "tRecall_75": 0.4, + "tRecall_90": 0.28125, + "token_acc": 0.939945652173913 + }, + { + "epoch": 0.9855888686433659, + "grad_norm": 0.7166591882705688, + "learning_rate": 5.570860703850711e-09, + "loss": 0.1591552734375, + "step": 2975, + "tIoU": 0.5079808712005616, + "tRecall_50": 0.5375, + "tRecall_75": 0.3875, + "tRecall_90": 0.21875, + "token_acc": 0.938113413304253 + }, + { + "epoch": 0.9872453205234388, + "grad_norm": 0.4753721356391907, + "learning_rate": 4.3768680241590425e-09, + "loss": 0.14058837890625, + "step": 2980, + "tIoU": 0.5579365253448486, + "tRecall_50": 0.6, + "tRecall_75": 0.43125, + "tRecall_90": 0.325, + "token_acc": 0.946195652173913 + }, + { + "epoch": 0.9889017724035116, + "grad_norm": 0.6353902220726013, + "learning_rate": 3.3266512910651303e-09, + "loss": 0.1630615234375, + "step": 2985, + "tIoU": 0.5626879692077636, + "tRecall_50": 0.60625, + "tRecall_75": 0.3875, + "tRecall_90": 0.24375, + "token_acc": 0.9439912996193583 + }, + { + "epoch": 0.9905582242835845, + "grad_norm": 0.7356439828872681, + "learning_rate": 2.4202407302093e-09, + "loss": 0.162890625, + "step": 2990, + "tIoU": 0.5033055305480957, + "tRecall_50": 0.5625, + "tRecall_75": 0.36875, + "tRecall_90": 0.20625, + "token_acc": 0.9392039258451472 + }, + { + "epoch": 0.9922146761636574, + "grad_norm": 0.4798955023288727, + "learning_rate": 1.657662428434792e-09, + "loss": 0.16402587890625, + "step": 2995, + "tIoU": 0.5145360469818115, + "tRecall_50": 0.55625, + "tRecall_75": 0.33125, + "tRecall_90": 0.25, + "token_acc": 0.9397820163487739 + }, + { + "epoch": 0.9938711280437303, + "grad_norm": 0.8724395632743835, + "learning_rate": 1.0389383330361436e-09, + "loss": 0.1508544921875, + "step": 3000, + "tIoU": 0.5828092575073243, + "tRecall_50": 0.65625, + "tRecall_75": 0.4375, + "tRecall_90": 0.2625, + "token_acc": 0.9477693144722524 + }, + { + "epoch": 0.9955275799238033, + "grad_norm": 0.4469767212867737, + "learning_rate": 5.64086251130247e-10, + "loss": 0.161175537109375, + "step": 3005, + "tIoU": 0.514407730102539, + "tRecall_50": 0.5, + "tRecall_75": 0.375, + "tRecall_90": 0.275, + "token_acc": 0.9394843962008141 + }, + { + "epoch": 0.997184031803876, + "grad_norm": 0.6539129614830017, + "learning_rate": 2.331198491417608e-10, + "loss": 0.178851318359375, + "step": 3010, + "tIoU": 0.4971726894378662, + "tRecall_50": 0.5375, + "tRecall_75": 0.35, + "tRecall_90": 0.19375, + "token_acc": 0.9323553382233088 + }, + { + "epoch": 0.998840483683949, + "grad_norm": 0.34182894229888916, + "learning_rate": 4.604865241064626e-11, + "loss": 0.175439453125, + "step": 3015, + "tIoU": 0.5114048480987549, + "tRecall_50": 0.54375, + "tRecall_75": 0.36875, + "tRecall_90": 0.26875, + "token_acc": 0.9353085077466703 + } + ], + "logging_steps": 5, + "max_steps": 3019, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 8.424853533524125e+19, + "train_batch_size": 4, + "trial_name": null, + "trial_params": null +} diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/training_args.bin b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..fed7d3707442dc9f81bf6f24137376094070fd12 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5d9b27722e827b51bc8191b35c4d479d230e61d017d6e28a663426f7c890d116 +size 9297 diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/video_preprocessor_config.json b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/video_preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..e32b1d90356f74457edbfca27aa1e121dab7cdc7 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/video_preprocessor_config.json @@ -0,0 +1,41 @@ +{ + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" +} diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/zero_to_fp32.py b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..5995d6e6f04e43b989587aa9022a3aef0c66d694 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019/zero_to_fp32.py @@ -0,0 +1,760 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: +# python zero_to_fp32.py . output_dir/ +# or +# python zero_to_fp32.py . output_dir/ --safe_serialization + +import argparse +import torch +import glob +import math +import os +import re +import gc +import json +import numpy as np +from tqdm import tqdm +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device, weights_only=False) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + total_files = len(files) + state_dicts = [] + for f in tqdm(files, desc='Loading checkpoint shards'): + state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if ZERO_STAGE not in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +class GatheredTensor: + """ + A pseudo tensor that collects partitioned weights. + It is more memory efficient when there are multiple groups. + """ + + def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape): + self.flat_groups = flat_groups + self.flat_groups_offset = flat_groups_offset + self.offset = offset + self.partitioned_numel = partitioned_numel + self.shape = shape + self.dtype = self.flat_groups[0][0].dtype + + def contiguous(self): + """ + Merge partitioned weights from flat_groups into a single tensor. + """ + end_idx = self.offset + self.partitioned_numel + world_size = len(self.flat_groups) + pad_flat_param_chunks = [] + + for rank_i in range(world_size): + # for each rank, we need to collect weights from related group/groups + flat_groups_at_rank_i = self.flat_groups[rank_i] + start_group_id = None + end_group_id = None + for group_id in range(len(self.flat_groups_offset)): + if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]: + start_group_id = group_id + if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]: + end_group_id = group_id + break + # collect weights from related group/groups + for group_id in range(start_group_id, end_group_id + 1): + flat_tensor = flat_groups_at_rank_i[group_id] + start_offset = self.offset - self.flat_groups_offset[group_id] + end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id] + pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset]) + + # collect weights from all ranks + pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0) + param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous() + return param + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size + + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]])) + for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'): + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # memory efficient tensor + tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape) + state_dict[name] = tensor + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def to_torch_tensor(state_dict, return_empty_tensor=False): + """ + Convert state_dict of GatheredTensor to torch tensor + """ + torch_state_dict = {} + converted_tensors = {} + for name, tensor in state_dict.items(): + tensor_id = id(tensor) + if tensor_id in converted_tensors: # shared tensors + shared_tensor = torch_state_dict[converted_tensors[tensor_id]] + torch_state_dict[name] = shared_tensor + else: + converted_tensors[tensor_id] = name + if return_empty_tensor: + torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype) + else: + torch_state_dict[name] = tensor.contiguous() + return torch_state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag=None, + exclude_frozen_parameters=False, + lazy_mode=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient. + Convert the pesduo tensor to torch tensor by ``.contiguous()`` + + Returns: + - pytorch ``state_dict`` + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + Note: the above usage may not work if your application doesn't have sufficient free CPU memory. + You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. Or you can load state_dict in lazy mode :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu + for name, lazy_tensor in state_dict.item(): + tensor = lazy_tensor.contiguous() # to cpu + print(name, tensor) + # del tensor to release memory if it no longer in use + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + if lazy_mode: + return state_dict + else: + return to_torch_tensor(state_dict) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, + output_dir, + max_shard_size="5GB", + safe_serialization=False, + tag=None, + exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_dir``: directory to the pytorch fp32 state_dict output files + - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB + - ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`). + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + # Dependency pre-check + if safe_serialization: + try: + from safetensors.torch import save_file + except ImportError: + print('If you want to use `safe_serialization`, please `pip install safetensors`') + raise + if max_shard_size is not None: + try: + from huggingface_hub import split_torch_state_dict_into_shards + except ImportError: + print('If you want to use `max_shard_size`, please `pip install huggingface_hub`') + raise + + # Convert zero checkpoint to state_dict + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag, + exclude_frozen_parameters, + lazy_mode=True) + + # Shard the model if it is too big. + weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin" + if max_shard_size is not None: + filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors") + # an memory-efficient approach for sharding + empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True) + state_dict_split = split_torch_state_dict_into_shards(empty_state_dict, + filename_pattern=filename_pattern, + max_shard_size=max_shard_size) + else: + from collections import namedtuple + StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"]) + state_dict_split = StateDictSplit(is_sharded=False, + filename_to_tensors={weights_name: list(state_dict.keys())}) + + # Save the model by shard + os.makedirs(output_dir, exist_ok=True) + filename_to_tensors = state_dict_split.filename_to_tensors.items() + for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"): + shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors} + shard_state_dict = to_torch_tensor(shard_state_dict) + output_path = os.path.join(output_dir, shard_file) + if safe_serialization: + save_file(shard_state_dict, output_path, metadata={"format": "pt"}) + else: + torch.save(shard_state_dict, output_path) + # release the memory of current shard + for tensor_name in list(shard_state_dict.keys()): + del state_dict[tensor_name] + del shard_state_dict[tensor_name] + del shard_state_dict + gc.collect() + + # Save index if sharded + if state_dict_split.is_sharded: + index = { + "metadata": state_dict_split.metadata, + "weight_map": state_dict_split.tensor_to_filename, + } + save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json" + save_index_file = os.path.join(output_dir, save_index_file) + with open(save_index_file, "w", encoding="utf-8") as f: + content = json.dumps(index, indent=2, sort_keys=True) + "\n" + f.write(content) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info("Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info("Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument("output_dir", + type=str, + help="directory to the pytorch fp32 state_dict output files" + "(e.g. path/checkpoint-12-output/)") + parser.add_argument( + "--max_shard_size", + type=str, + default="5GB", + help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size" + "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`" + "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances" + "without CPU OOM issues.") + parser.add_argument( + "--safe_serialization", + default=False, + action='store_true', + help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_dir, + max_shard_size=args.max_shard_size, + safe_serialization=args.safe_serialization, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/logging.jsonl b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/logging.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..587b7dc0bd4ef722292998f473185684e5a60b07 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/logging.jsonl @@ -0,0 +1,606 @@ +{"loss": 3.31054688, "grad_norm": 37.5207634, "learning_rate": 1.1e-07, "tIoU": 0.0, "tRecall_50": 0.0, "tRecall_75": 0.0, "tRecall_90": 0.0, "token_acc": 0.75857339, "epoch": 0.00033129, "global_step/max_steps": "1/3019", "percentage": "0.03%", "elapsed_time": "2m 55s", "remaining_time": "6d 3h 0m 8s", "memory(GiB)": 32.34, "train_speed(iter/s)": 0.005703} +{"loss": 3.28759766, "grad_norm": 33.39270401, "learning_rate": 5.5e-07, "tIoU": 0.0, "tRecall_50": 0.0, "tRecall_75": 0.0, "tRecall_90": 0.0, "token_acc": 0.75403087, "epoch": 0.00165645, "global_step/max_steps": "5/3019", "percentage": "0.17%", "elapsed_time": "4m 26s", "remaining_time": "1d 20h 39m 32s", "memory(GiB)": 38.54, "train_speed(iter/s)": 0.018747} +{"loss": 2.9546875, "grad_norm": 23.32330894, "learning_rate": 1.1e-06, "tIoU": 0.0, "tRecall_50": 0.0, "tRecall_75": 0.0, "tRecall_90": 0.0, "token_acc": 0.77903137, "epoch": 0.0033129, "global_step/max_steps": "10/3019", "percentage": "0.33%", "elapsed_time": "5m 55s", "remaining_time": "1d 5h 43m 0s", "memory(GiB)": 38.54, "train_speed(iter/s)": 0.028126} +{"loss": 1.83085937, "grad_norm": 15.74441528, "learning_rate": 1.65e-06, "tIoU": 0.0, "tRecall_50": 0.0, "tRecall_75": 0.0, "tRecall_90": 0.0, "token_acc": 0.82375479, "epoch": 0.00496936, "global_step/max_steps": "15/3019", "percentage": "0.50%", "elapsed_time": "7m 0s", "remaining_time": "23h 24m 49s", "memory(GiB)": 38.61, "train_speed(iter/s)": 0.035639} +{"loss": 0.6703125, "grad_norm": 13.48304749, "learning_rate": 2.2e-06, "tIoU": 0.15047524, "tRecall_50": 0.15, "tRecall_75": 0.09375, "tRecall_90": 0.05, "token_acc": 0.84282212, "epoch": 0.00662581, "global_step/max_steps": "20/3019", "percentage": "0.66%", "elapsed_time": "8m 5s", "remaining_time": "20h 14m 20s", "memory(GiB)": 38.61, "train_speed(iter/s)": 0.041161} +{"loss": 0.21407471, "grad_norm": 1.42586362, "learning_rate": 2.75e-06, "tIoU": 0.47949848, "tRecall_50": 0.5, "tRecall_75": 0.2875, "tRecall_90": 0.18125, "token_acc": 0.92793038, "epoch": 0.00828226, "global_step/max_steps": "25/3019", "percentage": "0.83%", "elapsed_time": "9m 9s", "remaining_time": "18h 17m 44s", "memory(GiB)": 38.61, "train_speed(iter/s)": 0.045457} +{"loss": 0.17580566, "grad_norm": 1.97630155, "learning_rate": 3.3e-06, "tIoU": 0.52486005, "tRecall_50": 0.59375, "tRecall_75": 0.35625, "tRecall_90": 0.2125, "token_acc": 0.93589744, "epoch": 0.00993871, "global_step/max_steps": "30/3019", "percentage": "0.99%", "elapsed_time": "10m 18s", "remaining_time": "17h 6m 36s", "memory(GiB)": 38.61, "train_speed(iter/s)": 0.048526} +{"loss": 0.19344482, "grad_norm": 2.23508668, "learning_rate": 3.85e-06, "tIoU": 0.48079414, "tRecall_50": 0.5125, "tRecall_75": 0.33125, "tRecall_90": 0.225, "token_acc": 0.92968111, "epoch": 0.01159516, "global_step/max_steps": "35/3019", "percentage": "1.16%", "elapsed_time": "11m 25s", "remaining_time": "16h 13m 55s", "memory(GiB)": 38.61, "train_speed(iter/s)": 0.051065} +{"loss": 0.19040527, "grad_norm": 1.30675423, "learning_rate": 4.4e-06, "tIoU": 0.45671964, "tRecall_50": 0.475, "tRecall_75": 0.275, "tRecall_90": 0.15625, "token_acc": 0.92921074, "epoch": 0.01325162, "global_step/max_steps": "40/3019", "percentage": "1.32%", "elapsed_time": "12m 32s", "remaining_time": "15h 33m 27s", "memory(GiB)": 38.61, "train_speed(iter/s)": 0.05319} +{"loss": 0.18565674, "grad_norm": 1.17370319, "learning_rate": 4.95e-06, "tIoU": 0.4578301, "tRecall_50": 0.5, "tRecall_75": 0.34375, "tRecall_90": 0.24375, "token_acc": 0.93601313, "epoch": 0.01490807, "global_step/max_steps": "45/3019", "percentage": "1.49%", "elapsed_time": "13m 33s", "remaining_time": "14h 56m 24s", "memory(GiB)": 38.61, "train_speed(iter/s)": 0.055295} +{"loss": 0.19178467, "grad_norm": 1.69406569, "learning_rate": 5.49e-06, "tIoU": 0.44655814, "tRecall_50": 0.44375, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.9292543, "epoch": 0.01656452, "global_step/max_steps": "50/3019", "percentage": "1.66%", "elapsed_time": "14m 41s", "remaining_time": "14h 32m 29s", "memory(GiB)": 38.67, "train_speed(iter/s)": 0.056715} +{"loss": 0.18096924, "grad_norm": 2.11928463, "learning_rate": 6.04e-06, "tIoU": 0.44701719, "tRecall_50": 0.5, "tRecall_75": 0.31875, "tRecall_90": 0.16875, "token_acc": 0.92798241, "epoch": 0.01822097, "global_step/max_steps": "55/3019", "percentage": "1.82%", "elapsed_time": "15m 46s", "remaining_time": "14h 10m 10s", "memory(GiB)": 38.67, "train_speed(iter/s)": 0.058106} +{"loss": 0.21026001, "grad_norm": 1.13456714, "learning_rate": 6.59e-06, "tIoU": 0.39896743, "tRecall_50": 0.4125, "tRecall_75": 0.30625, "tRecall_90": 0.19375, "token_acc": 0.92584515, "epoch": 0.01987742, "global_step/max_steps": "60/3019", "percentage": "1.99%", "elapsed_time": "16m 52s", "remaining_time": "13h 51m 56s", "memory(GiB)": 38.67, "train_speed(iter/s)": 0.059279} +{"loss": 0.19576416, "grad_norm": 2.07912922, "learning_rate": 7.14e-06, "tIoU": 0.45177331, "tRecall_50": 0.51875, "tRecall_75": 0.26875, "tRecall_90": 0.14375, "token_acc": 0.92733189, "epoch": 0.02153387, "global_step/max_steps": "65/3019", "percentage": "2.15%", "elapsed_time": "17m 56s", "remaining_time": "13h 35m 6s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.060401} +{"loss": 0.19819336, "grad_norm": 2.35622787, "learning_rate": 7.69e-06, "tIoU": 0.48453369, "tRecall_50": 0.53125, "tRecall_75": 0.35625, "tRecall_90": 0.19375, "token_acc": 0.92911668, "epoch": 0.02319033, "global_step/max_steps": "70/3019", "percentage": "2.32%", "elapsed_time": "19m 2s", "remaining_time": "13h 22m 5s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.061277} +{"loss": 0.18869629, "grad_norm": 0.95555055, "learning_rate": 8.24e-06, "tIoU": 0.44451766, "tRecall_50": 0.4875, "tRecall_75": 0.3375, "tRecall_90": 0.1625, "token_acc": 0.92720829, "epoch": 0.02484678, "global_step/max_steps": "75/3019", "percentage": "2.48%", "elapsed_time": "20m 7s", "remaining_time": "13h 10m 6s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.062102} +{"loss": 0.18936768, "grad_norm": 1.0119189, "learning_rate": 8.79e-06, "tIoU": 0.43130989, "tRecall_50": 0.46875, "tRecall_75": 0.25625, "tRecall_90": 0.15, "token_acc": 0.92785189, "epoch": 0.02650323, "global_step/max_steps": "80/3019", "percentage": "2.65%", "elapsed_time": "21m 14s", "remaining_time": "13h 0m 33s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.062754} +{"loss": 0.20290527, "grad_norm": 1.28616285, "learning_rate": 9.34e-06, "tIoU": 0.42988148, "tRecall_50": 0.44375, "tRecall_75": 0.3125, "tRecall_90": 0.19375, "token_acc": 0.9288451, "epoch": 0.02815968, "global_step/max_steps": "85/3019", "percentage": "2.82%", "elapsed_time": "22m 22s", "remaining_time": "12h 52m 22s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.063312} +{"loss": 0.18587646, "grad_norm": 0.95923662, "learning_rate": 9.89e-06, "tIoU": 0.48898201, "tRecall_50": 0.525, "tRecall_75": 0.36875, "tRecall_90": 0.1875, "token_acc": 0.93451569, "epoch": 0.02981613, "global_step/max_steps": "90/3019", "percentage": "2.98%", "elapsed_time": "23m 32s", "remaining_time": "12h 46m 18s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.063704} +{"loss": 0.18406982, "grad_norm": 1.2948693, "learning_rate": 1e-05, "tIoU": 0.44582033, "tRecall_50": 0.49375, "tRecall_75": 0.30625, "tRecall_90": 0.14375, "token_acc": 0.92777627, "epoch": 0.03147259, "global_step/max_steps": "95/3019", "percentage": "3.15%", "elapsed_time": "24m 40s", "remaining_time": "12h 39m 31s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.064163} +{"loss": 0.2097168, "grad_norm": 1.08886003, "learning_rate": 1e-05, "tIoU": 0.40095315, "tRecall_50": 0.425, "tRecall_75": 0.21875, "tRecall_90": 0.1375, "token_acc": 0.92263688, "epoch": 0.03312904, "global_step/max_steps": "100/3019", "percentage": "3.31%", "elapsed_time": "25m 47s", "remaining_time": "12h 32m 58s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.064611} +{"loss": 0.2105957, "grad_norm": 0.79802555, "learning_rate": 1e-05, "tIoU": 0.42629294, "tRecall_50": 0.4625, "tRecall_75": 0.275, "tRecall_90": 0.10625, "token_acc": 0.92286874, "epoch": 0.03478549, "global_step/max_steps": "105/3019", "percentage": "3.48%", "elapsed_time": "26m 54s", "remaining_time": "12h 26m 42s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.065041} +{"loss": 0.20194092, "grad_norm": 1.25252783, "learning_rate": 1e-05, "tIoU": 0.40687137, "tRecall_50": 0.4375, "tRecall_75": 0.2375, "tRecall_90": 0.14375, "token_acc": 0.92525143, "epoch": 0.03644194, "global_step/max_steps": "110/3019", "percentage": "3.64%", "elapsed_time": "27m 59s", "remaining_time": "12h 20m 10s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.065503} +{"loss": 0.20822754, "grad_norm": 1.621292, "learning_rate": 1e-05, "tIoU": 0.44732857, "tRecall_50": 0.50625, "tRecall_75": 0.2625, "tRecall_90": 0.14375, "token_acc": 0.92519042, "epoch": 0.03809839, "global_step/max_steps": "115/3019", "percentage": "3.81%", "elapsed_time": "29m 2s", "remaining_time": "12h 13m 32s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.065981} +{"loss": 0.20827637, "grad_norm": 0.85919839, "learning_rate": 1e-05, "tIoU": 0.41911063, "tRecall_50": 0.44375, "tRecall_75": 0.29375, "tRecall_90": 0.15625, "token_acc": 0.92253137, "epoch": 0.03975485, "global_step/max_steps": "120/3019", "percentage": "3.97%", "elapsed_time": "30m 9s", "remaining_time": "12h 8m 33s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.066318} +{"loss": 0.20395508, "grad_norm": 4.02284288, "learning_rate": 1e-05, "tIoU": 0.41852198, "tRecall_50": 0.4375, "tRecall_75": 0.26875, "tRecall_90": 0.14375, "token_acc": 0.92345277, "epoch": 0.0414113, "global_step/max_steps": "125/3019", "percentage": "4.14%", "elapsed_time": "31m 17s", "remaining_time": "12h 4m 22s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.066586} +{"loss": 0.22897949, "grad_norm": 1.01519144, "learning_rate": 1e-05, "tIoU": 0.32226543, "tRecall_50": 0.35, "tRecall_75": 0.21875, "tRecall_90": 0.125, "token_acc": 0.91905793, "epoch": 0.04306775, "global_step/max_steps": "130/3019", "percentage": "4.31%", "elapsed_time": "32m 22s", "remaining_time": "11h 59m 26s", "memory(GiB)": 39.41, "train_speed(iter/s)": 0.066926} +{"loss": 0.23083496, "grad_norm": 0.7622081, "learning_rate": 9.99e-06, "tIoU": 0.34840963, "tRecall_50": 0.375, "tRecall_75": 0.2, "tRecall_90": 0.10625, "token_acc": 0.91600764, "epoch": 0.0447242, "global_step/max_steps": "135/3019", "percentage": "4.47%", "elapsed_time": "33m 26s", "remaining_time": "11h 54m 20s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.067288} +{"loss": 0.21535645, "grad_norm": 1.06565773, "learning_rate": 9.99e-06, "tIoU": 0.34590023, "tRecall_50": 0.34375, "tRecall_75": 0.21875, "tRecall_90": 0.15625, "token_acc": 0.92095186, "epoch": 0.04638065, "global_step/max_steps": "140/3019", "percentage": "4.64%", "elapsed_time": "34m 30s", "remaining_time": "11h 49m 31s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.067627} +{"loss": 0.19082031, "grad_norm": 1.28969872, "learning_rate": 9.99e-06, "tIoU": 0.46801386, "tRecall_50": 0.5125, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.93227744, "epoch": 0.0480371, "global_step/max_steps": "145/3019", "percentage": "4.80%", "elapsed_time": "35m 37s", "remaining_time": "11h 46m 1s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.067845} +{"loss": 0.20524902, "grad_norm": 0.8637948, "learning_rate": 9.99e-06, "tIoU": 0.43006191, "tRecall_50": 0.43125, "tRecall_75": 0.2875, "tRecall_90": 0.14375, "token_acc": 0.92740175, "epoch": 0.04969356, "global_step/max_steps": "150/3019", "percentage": "4.97%", "elapsed_time": "36m 43s", "remaining_time": "11h 42m 29s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.068068} +{"loss": 0.1802124, "grad_norm": 1.60450292, "learning_rate": 9.99e-06, "tIoU": 0.50697365, "tRecall_50": 0.55625, "tRecall_75": 0.4, "tRecall_90": 0.21875, "token_acc": 0.93438606, "epoch": 0.05135001, "global_step/max_steps": "155/3019", "percentage": "5.13%", "elapsed_time": "37m 51s", "remaining_time": "11h 39m 34s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.068233} +{"loss": 0.19678955, "grad_norm": 1.54541063, "learning_rate": 9.99e-06, "tIoU": 0.44966755, "tRecall_50": 0.5, "tRecall_75": 0.3125, "tRecall_90": 0.16875, "token_acc": 0.9265506, "epoch": 0.05300646, "global_step/max_steps": "160/3019", "percentage": "5.30%", "elapsed_time": "38m 55s", "remaining_time": "11h 35m 26s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.068518} +{"loss": 0.22437134, "grad_norm": 0.83483607, "learning_rate": 9.98e-06, "tIoU": 0.44931426, "tRecall_50": 0.475, "tRecall_75": 0.30625, "tRecall_90": 0.1875, "token_acc": 0.9201304, "epoch": 0.05466291, "global_step/max_steps": "165/3019", "percentage": "5.47%", "elapsed_time": "40m 0s", "remaining_time": "11h 31m 59s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.068738} +{"loss": 0.19936523, "grad_norm": 1.08888984, "learning_rate": 9.98e-06, "tIoU": 0.43450875, "tRecall_50": 0.45625, "tRecall_75": 0.2625, "tRecall_90": 0.13125, "token_acc": 0.92548804, "epoch": 0.05631936, "global_step/max_steps": "170/3019", "percentage": "5.63%", "elapsed_time": "41m 5s", "remaining_time": "11h 28m 40s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.068949} +{"loss": 0.19187012, "grad_norm": 1.49407125, "learning_rate": 9.98e-06, "tIoU": 0.45506492, "tRecall_50": 0.49375, "tRecall_75": 0.3375, "tRecall_90": 0.15625, "token_acc": 0.92964276, "epoch": 0.05797582, "global_step/max_steps": "175/3019", "percentage": "5.80%", "elapsed_time": "42m 9s", "remaining_time": "11h 25m 7s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.069184} +{"loss": 0.20852051, "grad_norm": 0.89018607, "learning_rate": 9.98e-06, "tIoU": 0.37606673, "tRecall_50": 0.40625, "tRecall_75": 0.24375, "tRecall_90": 0.11875, "token_acc": 0.9240852, "epoch": 0.05963227, "global_step/max_steps": "180/3019", "percentage": "5.96%", "elapsed_time": "43m 16s", "remaining_time": "11h 22m 36s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.069318} +{"loss": 0.19017334, "grad_norm": 0.75073415, "learning_rate": 9.97e-06, "tIoU": 0.46780953, "tRecall_50": 0.525, "tRecall_75": 0.3375, "tRecall_90": 0.15, "token_acc": 0.92923497, "epoch": 0.06128872, "global_step/max_steps": "185/3019", "percentage": "6.13%", "elapsed_time": "44m 24s", "remaining_time": "11h 20m 16s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.069433} +{"loss": 0.2397583, "grad_norm": 1.24985814, "learning_rate": 9.97e-06, "tIoU": 0.3786418, "tRecall_50": 0.38125, "tRecall_75": 0.24375, "tRecall_90": 0.125, "token_acc": 0.91378841, "epoch": 0.06294517, "global_step/max_steps": "190/3019", "percentage": "6.29%", "elapsed_time": "45m 32s", "remaining_time": "11h 18m 9s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.069526} +{"loss": 0.19202271, "grad_norm": 1.1122005, "learning_rate": 9.97e-06, "tIoU": 0.460465, "tRecall_50": 0.46875, "tRecall_75": 0.3375, "tRecall_90": 0.2, "token_acc": 0.93313319, "epoch": 0.06460162, "global_step/max_steps": "195/3019", "percentage": "6.46%", "elapsed_time": "46m 36s", "remaining_time": "11h 15m 1s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.069725} +{"loss": 0.22727051, "grad_norm": 0.80761606, "learning_rate": 9.97e-06, "tIoU": 0.32807245, "tRecall_50": 0.35, "tRecall_75": 0.2, "tRecall_90": 0.14375, "token_acc": 0.9176663, "epoch": 0.06625808, "global_step/max_steps": "200/3019", "percentage": "6.62%", "elapsed_time": "47m 40s", "remaining_time": "11h 11m 51s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.06993} +{"loss": 0.20080566, "grad_norm": 1.58148253, "learning_rate": 9.96e-06, "tIoU": 0.41258435, "tRecall_50": 0.475, "tRecall_75": 0.2375, "tRecall_90": 0.15, "token_acc": 0.92710893, "epoch": 0.06791453, "global_step/max_steps": "205/3019", "percentage": "6.79%", "elapsed_time": "48m 45s", "remaining_time": "11h 9m 15s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.070077} +{"loss": 0.19127197, "grad_norm": 0.98811877, "learning_rate": 9.96e-06, "tIoU": 0.45179911, "tRecall_50": 0.5, "tRecall_75": 0.28125, "tRecall_90": 0.13125, "token_acc": 0.92661049, "epoch": 0.06957098, "global_step/max_steps": "210/3019", "percentage": "6.96%", "elapsed_time": "49m 52s", "remaining_time": "11h 7m 13s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.070166} +{"loss": 0.20115967, "grad_norm": 0.8627184, "learning_rate": 9.96e-06, "tIoU": 0.43936911, "tRecall_50": 0.44375, "tRecall_75": 0.26875, "tRecall_90": 0.15625, "token_acc": 0.92553191, "epoch": 0.07122743, "global_step/max_steps": "215/3019", "percentage": "7.12%", "elapsed_time": "50m 57s", "remaining_time": "11h 4m 36s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.070317} +{"loss": 0.20610352, "grad_norm": 0.8651228, "learning_rate": 9.95e-06, "tIoU": 0.41656847, "tRecall_50": 0.4125, "tRecall_75": 0.23125, "tRecall_90": 0.1375, "token_acc": 0.92570336, "epoch": 0.07288388, "global_step/max_steps": "220/3019", "percentage": "7.29%", "elapsed_time": "52m 1s", "remaining_time": "11h 1m 59s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.070469} +{"loss": 0.20802002, "grad_norm": 0.74430901, "learning_rate": 9.95e-06, "tIoU": 0.41522841, "tRecall_50": 0.425, "tRecall_75": 0.30625, "tRecall_90": 0.1625, "token_acc": 0.92487752, "epoch": 0.07454033, "global_step/max_steps": "225/3019", "percentage": "7.45%", "elapsed_time": "53m 5s", "remaining_time": "10h 59m 11s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.070642} +{"loss": 0.19639893, "grad_norm": 0.64825928, "learning_rate": 9.94e-06, "tIoU": 0.42443876, "tRecall_50": 0.45, "tRecall_75": 0.3, "tRecall_90": 0.1625, "token_acc": 0.92958904, "epoch": 0.07619679, "global_step/max_steps": "230/3019", "percentage": "7.62%", "elapsed_time": "54m 9s", "remaining_time": "10h 56m 41s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.070783} +{"loss": 0.1994873, "grad_norm": 0.86219352, "learning_rate": 9.94e-06, "tIoU": 0.43390598, "tRecall_50": 0.4625, "tRecall_75": 0.28125, "tRecall_90": 0.14375, "token_acc": 0.92818428, "epoch": 0.07785324, "global_step/max_steps": "235/3019", "percentage": "7.78%", "elapsed_time": "55m 12s", "remaining_time": "10h 54m 6s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.070937} +{"loss": 0.2104248, "grad_norm": 0.96819675, "learning_rate": 9.94e-06, "tIoU": 0.35329993, "tRecall_50": 0.35, "tRecall_75": 0.2, "tRecall_90": 0.1125, "token_acc": 0.92343451, "epoch": 0.07950969, "global_step/max_steps": "240/3019", "percentage": "7.95%", "elapsed_time": "56m 17s", "remaining_time": "10h 51m 49s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.071057} +{"loss": 0.20270996, "grad_norm": 2.02644372, "learning_rate": 9.93e-06, "tIoU": 0.40817695, "tRecall_50": 0.43125, "tRecall_75": 0.275, "tRecall_90": 0.2, "token_acc": 0.92486339, "epoch": 0.08116614, "global_step/max_steps": "245/3019", "percentage": "8.12%", "elapsed_time": "57m 22s", "remaining_time": "10h 49m 39s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.071166} +{"loss": 0.21035156, "grad_norm": 1.12672281, "learning_rate": 9.93e-06, "tIoU": 0.41781602, "tRecall_50": 0.4625, "tRecall_75": 0.2875, "tRecall_90": 0.1625, "token_acc": 0.92142857, "epoch": 0.08282259, "global_step/max_steps": "250/3019", "percentage": "8.28%", "elapsed_time": "58m 30s", "remaining_time": "10h 48m 5s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.07121} +{"loss": 0.20733643, "grad_norm": 0.79863852, "learning_rate": 9.92e-06, "tIoU": 0.45475945, "tRecall_50": 0.49375, "tRecall_75": 0.30625, "tRecall_90": 0.20625, "token_acc": 0.92694938, "epoch": 0.08447905, "global_step/max_steps": "255/3019", "percentage": "8.45%", "elapsed_time": "59m 36s", "remaining_time": "10h 46m 4s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.071302} +{"loss": 0.21896973, "grad_norm": 0.78025657, "learning_rate": 9.92e-06, "tIoU": 0.39062262, "tRecall_50": 0.43125, "tRecall_75": 0.225, "tRecall_90": 0.125, "token_acc": 0.92245119, "epoch": 0.0861355, "global_step/max_steps": "260/3019", "percentage": "8.61%", "elapsed_time": "1h 0m 46s", "remaining_time": "10h 44m 58s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.071294} +{"loss": 0.21519775, "grad_norm": 0.69445664, "learning_rate": 9.91e-06, "tIoU": 0.37258186, "tRecall_50": 0.40625, "tRecall_75": 0.23125, "tRecall_90": 0.1125, "token_acc": 0.91791655, "epoch": 0.08779195, "global_step/max_steps": "265/3019", "percentage": "8.78%", "elapsed_time": "1h 1m 54s", "remaining_time": "10h 43m 21s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.071345} +{"loss": 0.23343506, "grad_norm": 0.74982196, "learning_rate": 9.91e-06, "tIoU": 0.38320358, "tRecall_50": 0.39375, "tRecall_75": 0.24375, "tRecall_90": 0.1375, "token_acc": 0.91368479, "epoch": 0.0894484, "global_step/max_steps": "270/3019", "percentage": "8.94%", "elapsed_time": "1h 2m 59s", "remaining_time": "10h 41m 18s", "memory(GiB)": 39.42, "train_speed(iter/s)": 0.071442} +{"loss": 0.20540771, "grad_norm": 0.99642521, "learning_rate": 9.9e-06, "tIoU": 0.44122682, "tRecall_50": 0.5, "tRecall_75": 0.25, "tRecall_90": 0.1375, "token_acc": 0.92316076, "epoch": 0.09110485, "global_step/max_steps": "275/3019", "percentage": "9.11%", "elapsed_time": "1h 4m 6s", "remaining_time": "10h 39m 36s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.071501} +{"loss": 0.20683594, "grad_norm": 0.92143154, "learning_rate": 9.9e-06, "tIoU": 0.42021685, "tRecall_50": 0.4375, "tRecall_75": 0.29375, "tRecall_90": 0.1625, "token_acc": 0.92622951, "epoch": 0.09276131, "global_step/max_steps": "280/3019", "percentage": "9.27%", "elapsed_time": "1h 5m 8s", "remaining_time": "10h 37m 16s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.071633} +{"loss": 0.19711914, "grad_norm": 3.09095407, "learning_rate": 9.89e-06, "tIoU": 0.43810601, "tRecall_50": 0.48125, "tRecall_75": 0.2875, "tRecall_90": 0.16875, "token_acc": 0.92906053, "epoch": 0.09441776, "global_step/max_steps": "285/3019", "percentage": "9.44%", "elapsed_time": "1h 6m 14s", "remaining_time": "10h 35m 29s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.071702} +{"loss": 0.2024292, "grad_norm": 1.04584026, "learning_rate": 9.89e-06, "tIoU": 0.44785562, "tRecall_50": 0.5, "tRecall_75": 0.30625, "tRecall_90": 0.1875, "token_acc": 0.92991546, "epoch": 0.09607421, "global_step/max_steps": "290/3019", "percentage": "9.61%", "elapsed_time": "1h 7m 22s", "remaining_time": "10h 34m 5s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.071731} +{"loss": 0.20622559, "grad_norm": 0.84568727, "learning_rate": 9.88e-06, "tIoU": 0.42615447, "tRecall_50": 0.45, "tRecall_75": 0.3, "tRecall_90": 0.16875, "token_acc": 0.92471358, "epoch": 0.09773066, "global_step/max_steps": "295/3019", "percentage": "9.77%", "elapsed_time": "1h 8m 24s", "remaining_time": "10h 31m 42s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.071868} +{"loss": 0.20479736, "grad_norm": 0.96537691, "learning_rate": 9.87e-06, "tIoU": 0.39232655, "tRecall_50": 0.4125, "tRecall_75": 0.28125, "tRecall_90": 0.14375, "token_acc": 0.92420538, "epoch": 0.09938711, "global_step/max_steps": "300/3019", "percentage": "9.94%", "elapsed_time": "1h 9m 33s", "remaining_time": "10h 30m 21s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.071889} +{"loss": 0.19890137, "grad_norm": 0.83437806, "learning_rate": 9.87e-06, "tIoU": 0.41749234, "tRecall_50": 0.4375, "tRecall_75": 0.3125, "tRecall_90": 0.15, "token_acc": 0.93054024, "epoch": 0.10104356, "global_step/max_steps": "305/3019", "percentage": "10.10%", "elapsed_time": "1h 10m 38s", "remaining_time": "10h 28m 38s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.071954} +{"loss": 0.19197998, "grad_norm": 0.81543261, "learning_rate": 9.86e-06, "tIoU": 0.47261028, "tRecall_50": 0.55, "tRecall_75": 0.3125, "tRecall_90": 0.15, "token_acc": 0.9284159, "epoch": 0.10270002, "global_step/max_steps": "310/3019", "percentage": "10.27%", "elapsed_time": "1h 11m 46s", "remaining_time": "10h 27m 10s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.071989} +{"loss": 0.20172119, "grad_norm": 0.61932302, "learning_rate": 9.86e-06, "tIoU": 0.44944382, "tRecall_50": 0.49375, "tRecall_75": 0.3375, "tRecall_90": 0.1875, "token_acc": 0.92783505, "epoch": 0.10435647, "global_step/max_steps": "315/3019", "percentage": "10.43%", "elapsed_time": "1h 12m 51s", "remaining_time": "10h 25m 29s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.07205} +{"loss": 0.20203857, "grad_norm": 0.70888424, "learning_rate": 9.85e-06, "tIoU": 0.43042378, "tRecall_50": 0.45625, "tRecall_75": 0.29375, "tRecall_90": 0.15625, "token_acc": 0.92577488, "epoch": 0.10601292, "global_step/max_steps": "320/3019", "percentage": "10.60%", "elapsed_time": "1h 13m 55s", "remaining_time": "10h 23m 28s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072149} +{"loss": 0.196875, "grad_norm": 0.64781314, "learning_rate": 9.84e-06, "tIoU": 0.45046268, "tRecall_50": 0.5, "tRecall_75": 0.3375, "tRecall_90": 0.15625, "token_acc": 0.92847393, "epoch": 0.10766937, "global_step/max_steps": "325/3019", "percentage": "10.77%", "elapsed_time": "1h 15m 0s", "remaining_time": "10h 21m 43s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072218} +{"loss": 0.20408936, "grad_norm": 0.55446059, "learning_rate": 9.84e-06, "tIoU": 0.47177978, "tRecall_50": 0.5, "tRecall_75": 0.33125, "tRecall_90": 0.19375, "token_acc": 0.92755991, "epoch": 0.10932582, "global_step/max_steps": "330/3019", "percentage": "10.93%", "elapsed_time": "1h 16m 11s", "remaining_time": "10h 20m 51s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072184} +{"loss": 0.20200195, "grad_norm": 0.75530255, "learning_rate": 9.83e-06, "tIoU": 0.42468314, "tRecall_50": 0.45, "tRecall_75": 0.25625, "tRecall_90": 0.1375, "token_acc": 0.92318169, "epoch": 0.11098228, "global_step/max_steps": "335/3019", "percentage": "11.10%", "elapsed_time": "1h 17m 18s", "remaining_time": "10h 19m 20s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072227} +{"loss": 0.32507324, "grad_norm": 24.64958191, "learning_rate": 9.82e-06, "tIoU": 0.33823087, "tRecall_50": 0.3625, "tRecall_75": 0.23125, "tRecall_90": 0.16875, "token_acc": 0.91766304, "epoch": 0.11263873, "global_step/max_steps": "340/3019", "percentage": "11.26%", "elapsed_time": "1h 18m 26s", "remaining_time": "10h 18m 4s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072241} +{"loss": 0.1960083, "grad_norm": 1.22218299, "learning_rate": 9.82e-06, "tIoU": 0.46476555, "tRecall_50": 0.51875, "tRecall_75": 0.34375, "tRecall_90": 0.18125, "token_acc": 0.93188602, "epoch": 0.11429518, "global_step/max_steps": "345/3019", "percentage": "11.43%", "elapsed_time": "1h 19m 31s", "remaining_time": "10h 16m 19s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072311} +{"loss": 0.19267578, "grad_norm": 0.51911974, "learning_rate": 9.81e-06, "tIoU": 0.44423165, "tRecall_50": 0.49375, "tRecall_75": 0.28125, "tRecall_90": 0.1375, "token_acc": 0.9275834, "epoch": 0.11595163, "global_step/max_steps": "350/3019", "percentage": "11.59%", "elapsed_time": "1h 20m 36s", "remaining_time": "10h 14m 42s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072366} +{"loss": 0.19835205, "grad_norm": 0.77918869, "learning_rate": 9.8e-06, "tIoU": 0.43950558, "tRecall_50": 0.425, "tRecall_75": 0.2875, "tRecall_90": 0.15, "token_acc": 0.92437892, "epoch": 0.11760808, "global_step/max_steps": "355/3019", "percentage": "11.76%", "elapsed_time": "1h 21m 41s", "remaining_time": "10h 13m 0s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.07243} +{"loss": 0.2083252, "grad_norm": 0.64465159, "learning_rate": 9.79e-06, "tIoU": 0.43174319, "tRecall_50": 0.45625, "tRecall_75": 0.25625, "tRecall_90": 0.15625, "token_acc": 0.92476915, "epoch": 0.11926454, "global_step/max_steps": "360/3019", "percentage": "11.92%", "elapsed_time": "1h 22m 46s", "remaining_time": "10h 11m 26s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.07248} +{"loss": 0.19989014, "grad_norm": 0.64630592, "learning_rate": 9.79e-06, "tIoU": 0.40005889, "tRecall_50": 0.43125, "tRecall_75": 0.21875, "tRecall_90": 0.125, "token_acc": 0.92706911, "epoch": 0.12092099, "global_step/max_steps": "365/3019", "percentage": "12.09%", "elapsed_time": "1h 23m 53s", "remaining_time": "10h 10m 1s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.07251} +{"loss": 0.20377197, "grad_norm": 0.86979145, "learning_rate": 9.78e-06, "tIoU": 0.3834753, "tRecall_50": 0.4, "tRecall_75": 0.26875, "tRecall_90": 0.1375, "token_acc": 0.92379131, "epoch": 0.12257744, "global_step/max_steps": "370/3019", "percentage": "12.26%", "elapsed_time": "1h 25m 0s", "remaining_time": "10h 8m 33s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072549} +{"loss": 0.19049072, "grad_norm": 0.82842827, "learning_rate": 9.77e-06, "tIoU": 0.41577611, "tRecall_50": 0.43125, "tRecall_75": 0.30625, "tRecall_90": 0.2125, "token_acc": 0.93311582, "epoch": 0.12423389, "global_step/max_steps": "375/3019", "percentage": "12.42%", "elapsed_time": "1h 26m 4s", "remaining_time": "10h 6m 54s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072609} +{"loss": 0.19884033, "grad_norm": 0.54440016, "learning_rate": 9.76e-06, "tIoU": 0.4304636, "tRecall_50": 0.48125, "tRecall_75": 0.275, "tRecall_90": 0.1625, "token_acc": 0.9262698, "epoch": 0.12589034, "global_step/max_steps": "380/3019", "percentage": "12.59%", "elapsed_time": "1h 27m 6s", "remaining_time": "10h 4m 58s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072702} +{"loss": 0.196875, "grad_norm": 0.77499247, "learning_rate": 9.75e-06, "tIoU": 0.45472007, "tRecall_50": 0.4625, "tRecall_75": 0.3, "tRecall_90": 0.18125, "token_acc": 0.92814371, "epoch": 0.12754679, "global_step/max_steps": "385/3019", "percentage": "12.75%", "elapsed_time": "1h 28m 10s", "remaining_time": "10h 3m 15s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072772} +{"loss": 0.18959961, "grad_norm": 0.82476103, "learning_rate": 9.74e-06, "tIoU": 0.46105518, "tRecall_50": 0.45, "tRecall_75": 0.33125, "tRecall_90": 0.16875, "token_acc": 0.93123288, "epoch": 0.12920325, "global_step/max_steps": "390/3019", "percentage": "12.92%", "elapsed_time": "1h 29m 14s", "remaining_time": "10h 1m 37s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.07283} +{"loss": 0.20187988, "grad_norm": 1.26711059, "learning_rate": 9.74e-06, "tIoU": 0.41598783, "tRecall_50": 0.4625, "tRecall_75": 0.2625, "tRecall_90": 0.125, "token_acc": 0.92661049, "epoch": 0.1308597, "global_step/max_steps": "395/3019", "percentage": "13.08%", "elapsed_time": "1h 30m 21s", "remaining_time": "10h 0m 13s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072861} +{"loss": 0.20133057, "grad_norm": 0.72217333, "learning_rate": 9.73e-06, "tIoU": 0.43442955, "tRecall_50": 0.43125, "tRecall_75": 0.275, "tRecall_90": 0.1625, "token_acc": 0.92192192, "epoch": 0.13251615, "global_step/max_steps": "400/3019", "percentage": "13.25%", "elapsed_time": "1h 31m 26s", "remaining_time": "9h 58m 40s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.07291} +{"loss": 0.19154053, "grad_norm": 0.80751461, "learning_rate": 9.72e-06, "tIoU": 0.47398047, "tRecall_50": 0.53125, "tRecall_75": 0.31875, "tRecall_90": 0.14375, "token_acc": 0.93137255, "epoch": 0.1341726, "global_step/max_steps": "405/3019", "percentage": "13.42%", "elapsed_time": "1h 32m 30s", "remaining_time": "9h 57m 6s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072963} +{"loss": 0.18717041, "grad_norm": 0.66282046, "learning_rate": 9.71e-06, "tIoU": 0.48113313, "tRecall_50": 0.51875, "tRecall_75": 0.33125, "tRecall_90": 0.20625, "token_acc": 0.93194862, "epoch": 0.13582905, "global_step/max_steps": "410/3019", "percentage": "13.58%", "elapsed_time": "1h 33m 36s", "remaining_time": "9h 55m 42s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.072994} +{"loss": 0.20632324, "grad_norm": 0.70934844, "learning_rate": 9.7e-06, "tIoU": 0.41611781, "tRecall_50": 0.4625, "tRecall_75": 0.275, "tRecall_90": 0.125, "token_acc": 0.92209207, "epoch": 0.13748551, "global_step/max_steps": "415/3019", "percentage": "13.75%", "elapsed_time": "1h 34m 40s", "remaining_time": "9h 54m 3s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073057} +{"loss": 0.18984375, "grad_norm": 0.51630658, "learning_rate": 9.69e-06, "tIoU": 0.45911922, "tRecall_50": 0.4875, "tRecall_75": 0.30625, "tRecall_90": 0.19375, "token_acc": 0.93121403, "epoch": 0.13914196, "global_step/max_steps": "420/3019", "percentage": "13.91%", "elapsed_time": "1h 35m 48s", "remaining_time": "9h 52m 49s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073069} +{"loss": 0.19431152, "grad_norm": 0.64796185, "learning_rate": 9.68e-06, "tIoU": 0.41985331, "tRecall_50": 0.45625, "tRecall_75": 0.2375, "tRecall_90": 0.1125, "token_acc": 0.92468733, "epoch": 0.14079841, "global_step/max_steps": "425/3019", "percentage": "14.08%", "elapsed_time": "1h 36m 56s", "remaining_time": "9h 51m 40s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073069} +{"loss": 0.19625244, "grad_norm": 1.39156556, "learning_rate": 9.67e-06, "tIoU": 0.4425868, "tRecall_50": 0.46875, "tRecall_75": 0.24375, "tRecall_90": 0.125, "token_acc": 0.92717391, "epoch": 0.14245486, "global_step/max_steps": "430/3019", "percentage": "14.24%", "elapsed_time": "1h 38m 6s", "remaining_time": "9h 50m 40s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073053} +{"loss": 0.19682617, "grad_norm": 0.87462795, "learning_rate": 9.66e-06, "tIoU": 0.44972887, "tRecall_50": 0.45, "tRecall_75": 0.31875, "tRecall_90": 0.1625, "token_acc": 0.92726777, "epoch": 0.14411131, "global_step/max_steps": "435/3019", "percentage": "14.41%", "elapsed_time": "1h 39m 10s", "remaining_time": "9h 49m 5s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073107} +{"loss": 0.21339111, "grad_norm": 0.64941788, "learning_rate": 9.65e-06, "tIoU": 0.37221119, "tRecall_50": 0.36875, "tRecall_75": 0.225, "tRecall_90": 0.15, "token_acc": 0.92207084, "epoch": 0.14576777, "global_step/max_steps": "440/3019", "percentage": "14.57%", "elapsed_time": "1h 40m 18s", "remaining_time": "9h 47m 54s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073112} +{"loss": 0.18795166, "grad_norm": 0.63572156, "learning_rate": 9.64e-06, "tIoU": 0.46097903, "tRecall_50": 0.50625, "tRecall_75": 0.30625, "tRecall_90": 0.16875, "token_acc": 0.9316888, "epoch": 0.14742422, "global_step/max_steps": "445/3019", "percentage": "14.74%", "elapsed_time": "1h 41m 25s", "remaining_time": "9h 46m 39s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073126} +{"loss": 0.19854736, "grad_norm": 0.91340148, "learning_rate": 9.63e-06, "tIoU": 0.44006886, "tRecall_50": 0.5, "tRecall_75": 0.28125, "tRecall_90": 0.15, "token_acc": 0.92718844, "epoch": 0.14908067, "global_step/max_steps": "450/3019", "percentage": "14.91%", "elapsed_time": "1h 42m 28s", "remaining_time": "9h 44m 59s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073192} +{"loss": 0.18902588, "grad_norm": 1.31812131, "learning_rate": 9.62e-06, "tIoU": 0.48061619, "tRecall_50": 0.55, "tRecall_75": 0.31875, "tRecall_90": 0.1875, "token_acc": 0.93099671, "epoch": 0.15073712, "global_step/max_steps": "455/3019", "percentage": "15.07%", "elapsed_time": "1h 43m 36s", "remaining_time": "9h 43m 52s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073189} +{"loss": 0.20053711, "grad_norm": 0.89351892, "learning_rate": 9.61e-06, "tIoU": 0.43247547, "tRecall_50": 0.45625, "tRecall_75": 0.24375, "tRecall_90": 0.1375, "token_acc": 0.9250477, "epoch": 0.15239357, "global_step/max_steps": "460/3019", "percentage": "15.24%", "elapsed_time": "1h 44m 42s", "remaining_time": "9h 42m 30s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073217} +{"loss": 0.2102417, "grad_norm": 0.82909322, "learning_rate": 9.6e-06, "tIoU": 0.44379296, "tRecall_50": 0.4625, "tRecall_75": 0.275, "tRecall_90": 0.1875, "token_acc": 0.92622951, "epoch": 0.15405002, "global_step/max_steps": "465/3019", "percentage": "15.40%", "elapsed_time": "1h 45m 49s", "remaining_time": "9h 41m 16s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.07323} +{"loss": 0.19241943, "grad_norm": 0.70945978, "learning_rate": 9.59e-06, "tIoU": 0.48861909, "tRecall_50": 0.53125, "tRecall_75": 0.30625, "tRecall_90": 0.19375, "token_acc": 0.93124147, "epoch": 0.15570648, "global_step/max_steps": "470/3019", "percentage": "15.57%", "elapsed_time": "1h 46m 55s", "remaining_time": "9h 39m 54s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.07326} +{"loss": 0.19324951, "grad_norm": 0.50244701, "learning_rate": 9.58e-06, "tIoU": 0.43886647, "tRecall_50": 0.4875, "tRecall_75": 0.30625, "tRecall_90": 0.1875, "token_acc": 0.93175639, "epoch": 0.15736293, "global_step/max_steps": "475/3019", "percentage": "15.73%", "elapsed_time": "1h 48m 0s", "remaining_time": "9h 38m 27s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073298} +{"loss": 0.21569824, "grad_norm": 0.75572062, "learning_rate": 9.57e-06, "tIoU": 0.40017114, "tRecall_50": 0.43125, "tRecall_75": 0.25625, "tRecall_90": 0.1375, "token_acc": 0.92477514, "epoch": 0.15901938, "global_step/max_steps": "480/3019", "percentage": "15.90%", "elapsed_time": "1h 49m 5s", "remaining_time": "9h 37m 4s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.07333} +{"loss": 0.18979492, "grad_norm": 0.82273751, "learning_rate": 9.56e-06, "tIoU": 0.47867346, "tRecall_50": 0.50625, "tRecall_75": 0.3375, "tRecall_90": 0.16875, "token_acc": 0.93102508, "epoch": 0.16067583, "global_step/max_steps": "485/3019", "percentage": "16.06%", "elapsed_time": "1h 50m 10s", "remaining_time": "9h 35m 36s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073372} +{"loss": 0.19400635, "grad_norm": 0.92036211, "learning_rate": 9.55e-06, "tIoU": 0.4759007, "tRecall_50": 0.5375, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.93074416, "epoch": 0.16233228, "global_step/max_steps": "490/3019", "percentage": "16.23%", "elapsed_time": "1h 51m 12s", "remaining_time": "9h 34m 0s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073431} +{"loss": 0.18093262, "grad_norm": 0.84497905, "learning_rate": 9.54e-06, "tIoU": 0.47850103, "tRecall_50": 0.5, "tRecall_75": 0.34375, "tRecall_90": 0.20625, "token_acc": 0.9327342, "epoch": 0.16398874, "global_step/max_steps": "495/3019", "percentage": "16.40%", "elapsed_time": "1h 52m 16s", "remaining_time": "9h 32m 29s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073481} +{"loss": 0.21148682, "grad_norm": 0.77489555, "learning_rate": 9.53e-06, "tIoU": 0.43302107, "tRecall_50": 0.45, "tRecall_75": 0.28125, "tRecall_90": 0.175, "token_acc": 0.92219804, "epoch": 0.16564519, "global_step/max_steps": "500/3019", "percentage": "16.56%", "elapsed_time": "1h 53m 21s", "remaining_time": "9h 31m 7s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073511} +{"loss": 0.197229, "grad_norm": 0.64767849, "learning_rate": 9.51e-06, "tIoU": 0.43325477, "tRecall_50": 0.46875, "tRecall_75": 0.28125, "tRecall_90": 0.1375, "token_acc": 0.9282988, "epoch": 0.16730164, "global_step/max_steps": "505/3019", "percentage": "16.73%", "elapsed_time": "1h 54m 25s", "remaining_time": "9h 29m 40s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073551} +{"loss": 0.19720459, "grad_norm": 0.73413229, "learning_rate": 9.5e-06, "tIoU": 0.47381792, "tRecall_50": 0.4875, "tRecall_75": 0.31875, "tRecall_90": 0.19375, "token_acc": 0.93329718, "epoch": 0.16895809, "global_step/max_steps": "510/3019", "percentage": "16.89%", "elapsed_time": "1h 55m 32s", "remaining_time": "9h 28m 26s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073563} +{"loss": 0.19243774, "grad_norm": 0.94317436, "learning_rate": 9.49e-06, "tIoU": 0.40606904, "tRecall_50": 0.43125, "tRecall_75": 0.2875, "tRecall_90": 0.15625, "token_acc": 0.92718314, "epoch": 0.17061454, "global_step/max_steps": "515/3019", "percentage": "17.06%", "elapsed_time": "1h 56m 34s", "remaining_time": "9h 26m 47s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073632} +{"loss": 0.19864502, "grad_norm": 0.937365, "learning_rate": 9.48e-06, "tIoU": 0.45730491, "tRecall_50": 0.4875, "tRecall_75": 0.325, "tRecall_90": 0.16875, "token_acc": 0.93183053, "epoch": 0.172271, "global_step/max_steps": "520/3019", "percentage": "17.22%", "elapsed_time": "1h 57m 42s", "remaining_time": "9h 25m 39s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073631} +{"loss": 0.19825439, "grad_norm": 0.69108719, "learning_rate": 9.47e-06, "tIoU": 0.44774075, "tRecall_50": 0.45625, "tRecall_75": 0.33125, "tRecall_90": 0.16875, "token_acc": 0.93145051, "epoch": 0.17392745, "global_step/max_steps": "525/3019", "percentage": "17.39%", "elapsed_time": "1h 58m 46s", "remaining_time": "9h 24m 14s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073669} +{"loss": 0.19985352, "grad_norm": 0.49583554, "learning_rate": 9.46e-06, "tIoU": 0.43484712, "tRecall_50": 0.45625, "tRecall_75": 0.3, "tRecall_90": 0.19375, "token_acc": 0.92497961, "epoch": 0.1755839, "global_step/max_steps": "530/3019", "percentage": "17.56%", "elapsed_time": "1h 59m 53s", "remaining_time": "9h 23m 4s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073673} +{"loss": 0.19815674, "grad_norm": 0.58157647, "learning_rate": 9.44e-06, "tIoU": 0.4181746, "tRecall_50": 0.44375, "tRecall_75": 0.2625, "tRecall_90": 0.15, "token_acc": 0.92517007, "epoch": 0.17724035, "global_step/max_steps": "535/3019", "percentage": "17.72%", "elapsed_time": "2h 1m 0s", "remaining_time": "9h 21m 51s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073685} +{"loss": 0.18758545, "grad_norm": 0.83415341, "learning_rate": 9.43e-06, "tIoU": 0.41258726, "tRecall_50": 0.45625, "tRecall_75": 0.28125, "tRecall_90": 0.15625, "token_acc": 0.92573449, "epoch": 0.1788968, "global_step/max_steps": "540/3019", "percentage": "17.89%", "elapsed_time": "2h 2m 2s", "remaining_time": "9h 20m 17s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073741} +{"loss": 0.17341309, "grad_norm": 0.49157268, "learning_rate": 9.42e-06, "tIoU": 0.49922776, "tRecall_50": 0.55, "tRecall_75": 0.3875, "tRecall_90": 0.2375, "token_acc": 0.93624525, "epoch": 0.18055325, "global_step/max_steps": "545/3019", "percentage": "18.05%", "elapsed_time": "2h 3m 6s", "remaining_time": "9h 18m 49s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073785} +{"loss": 0.20540771, "grad_norm": 0.8604576, "learning_rate": 9.41e-06, "tIoU": 0.40863843, "tRecall_50": 0.4375, "tRecall_75": 0.28125, "tRecall_90": 0.1875, "token_acc": 0.92712329, "epoch": 0.18220971, "global_step/max_steps": "550/3019", "percentage": "18.22%", "elapsed_time": "2h 4m 16s", "remaining_time": "9h 17m 50s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073766} +{"loss": 0.1869751, "grad_norm": 0.8063392, "learning_rate": 9.39e-06, "tIoU": 0.47340059, "tRecall_50": 0.53125, "tRecall_75": 0.3375, "tRecall_90": 0.1875, "token_acc": 0.92735746, "epoch": 0.18386616, "global_step/max_steps": "555/3019", "percentage": "18.38%", "elapsed_time": "2h 5m 23s", "remaining_time": "9h 16m 41s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073769} +{"loss": 0.19299316, "grad_norm": 0.71959329, "learning_rate": 9.38e-06, "tIoU": 0.47695556, "tRecall_50": 0.49375, "tRecall_75": 0.35, "tRecall_90": 0.23125, "token_acc": 0.92843537, "epoch": 0.18552261, "global_step/max_steps": "560/3019", "percentage": "18.55%", "elapsed_time": "2h 6m 28s", "remaining_time": "9h 15m 20s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073799} +{"loss": 0.21213379, "grad_norm": 0.68721145, "learning_rate": 9.37e-06, "tIoU": 0.39610956, "tRecall_50": 0.4375, "tRecall_75": 0.25625, "tRecall_90": 0.13125, "token_acc": 0.92612859, "epoch": 0.18717906, "global_step/max_steps": "565/3019", "percentage": "18.71%", "elapsed_time": "2h 7m 36s", "remaining_time": "9h 14m 15s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073792} +{"loss": 0.18876953, "grad_norm": 0.65291715, "learning_rate": 9.35e-06, "tIoU": 0.44217601, "tRecall_50": 0.475, "tRecall_75": 0.3125, "tRecall_90": 0.2125, "token_acc": 0.93331504, "epoch": 0.18883551, "global_step/max_steps": "570/3019", "percentage": "18.88%", "elapsed_time": "2h 8m 43s", "remaining_time": "9h 13m 4s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.0738} +{"loss": 0.21179199, "grad_norm": 0.94554836, "learning_rate": 9.34e-06, "tIoU": 0.41131225, "tRecall_50": 0.4375, "tRecall_75": 0.29375, "tRecall_90": 0.1625, "token_acc": 0.92290929, "epoch": 0.19049197, "global_step/max_steps": "575/3019", "percentage": "19.05%", "elapsed_time": "2h 9m 53s", "remaining_time": "9h 12m 5s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073779} +{"loss": 0.20238037, "grad_norm": 0.62948179, "learning_rate": 9.33e-06, "tIoU": 0.42276864, "tRecall_50": 0.43125, "tRecall_75": 0.2875, "tRecall_90": 0.16875, "token_acc": 0.92471358, "epoch": 0.19214842, "global_step/max_steps": "580/3019", "percentage": "19.21%", "elapsed_time": "2h 11m 1s", "remaining_time": "9h 10m 59s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073775} +{"loss": 0.19185791, "grad_norm": 0.49901363, "learning_rate": 9.31e-06, "tIoU": 0.42323399, "tRecall_50": 0.46875, "tRecall_75": 0.3125, "tRecall_90": 0.1375, "token_acc": 0.92464635, "epoch": 0.19380487, "global_step/max_steps": "585/3019", "percentage": "19.38%", "elapsed_time": "2h 12m 7s", "remaining_time": "9h 9m 45s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073789} +{"loss": 0.19619141, "grad_norm": 0.65577042, "learning_rate": 9.3e-06, "tIoU": 0.44829621, "tRecall_50": 0.5125, "tRecall_75": 0.275, "tRecall_90": 0.16875, "token_acc": 0.92945037, "epoch": 0.19546132, "global_step/max_steps": "590/3019", "percentage": "19.54%", "elapsed_time": "2h 13m 15s", "remaining_time": "9h 8m 38s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073787} +{"loss": 0.20632324, "grad_norm": 1.52394342, "learning_rate": 9.29e-06, "tIoU": 0.46203609, "tRecall_50": 0.5125, "tRecall_75": 0.325, "tRecall_90": 0.16875, "token_acc": 0.92665572, "epoch": 0.19711777, "global_step/max_steps": "595/3019", "percentage": "19.71%", "elapsed_time": "2h 14m 27s", "remaining_time": "9h 7m 45s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073756} +{"loss": 0.20166016, "grad_norm": 0.68989497, "learning_rate": 9.27e-06, "tIoU": 0.44564409, "tRecall_50": 0.49375, "tRecall_75": 0.2875, "tRecall_90": 0.15625, "token_acc": 0.92647059, "epoch": 0.19877423, "global_step/max_steps": "600/3019", "percentage": "19.87%", "elapsed_time": "2h 15m 41s", "remaining_time": "9h 7m 3s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073698} +{"loss": 0.19608154, "grad_norm": 0.86281151, "learning_rate": 9.26e-06, "tIoU": 0.41976085, "tRecall_50": 0.425, "tRecall_75": 0.29375, "tRecall_90": 0.16875, "token_acc": 0.93057446, "epoch": 0.20043068, "global_step/max_steps": "605/3019", "percentage": "20.04%", "elapsed_time": "2h 16m 49s", "remaining_time": "9h 5m 56s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073696} +{"loss": 0.2102417, "grad_norm": 0.50180811, "learning_rate": 9.24e-06, "tIoU": 0.43528981, "tRecall_50": 0.44375, "tRecall_75": 0.275, "tRecall_90": 0.15625, "token_acc": 0.92447212, "epoch": 0.20208713, "global_step/max_steps": "610/3019", "percentage": "20.21%", "elapsed_time": "2h 18m 1s", "remaining_time": "9h 5m 3s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073662} +{"loss": 0.1876709, "grad_norm": 0.58079731, "learning_rate": 9.23e-06, "tIoU": 0.4596035, "tRecall_50": 0.475, "tRecall_75": 0.3125, "tRecall_90": 0.19375, "token_acc": 0.93118514, "epoch": 0.20374358, "global_step/max_steps": "615/3019", "percentage": "20.37%", "elapsed_time": "2h 19m 9s", "remaining_time": "9h 3m 56s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073659} +{"loss": 0.20012207, "grad_norm": 0.82113796, "learning_rate": 9.22e-06, "tIoU": 0.47874384, "tRecall_50": 0.4875, "tRecall_75": 0.325, "tRecall_90": 0.16875, "token_acc": 0.92888165, "epoch": 0.20540003, "global_step/max_steps": "620/3019", "percentage": "20.54%", "elapsed_time": "2h 20m 17s", "remaining_time": "9h 2m 49s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073658} +{"loss": 0.19661865, "grad_norm": 0.72327697, "learning_rate": 9.2e-06, "tIoU": 0.46714573, "tRecall_50": 0.5, "tRecall_75": 0.3375, "tRecall_90": 0.16875, "token_acc": 0.93034014, "epoch": 0.20705649, "global_step/max_steps": "625/3019", "percentage": "20.70%", "elapsed_time": "2h 21m 23s", "remaining_time": "9h 1m 37s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073668} +{"loss": 0.20330811, "grad_norm": 0.77745295, "learning_rate": 9.19e-06, "tIoU": 0.44649844, "tRecall_50": 0.46875, "tRecall_75": 0.30625, "tRecall_90": 0.15, "token_acc": 0.92498636, "epoch": 0.20871294, "global_step/max_steps": "630/3019", "percentage": "20.87%", "elapsed_time": "2h 22m 26s", "remaining_time": "9h 0m 7s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073716} +{"loss": 0.18858643, "grad_norm": 0.56142563, "learning_rate": 9.17e-06, "tIoU": 0.44100623, "tRecall_50": 0.48125, "tRecall_75": 0.29375, "tRecall_90": 0.15625, "token_acc": 0.92712329, "epoch": 0.21036939, "global_step/max_steps": "635/3019", "percentage": "21.03%", "elapsed_time": "2h 23m 30s", "remaining_time": "8h 58m 46s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073748} +{"loss": 0.19208984, "grad_norm": 0.55320626, "learning_rate": 9.16e-06, "tIoU": 0.44082904, "tRecall_50": 0.48125, "tRecall_75": 0.2625, "tRecall_90": 0.175, "token_acc": 0.92477514, "epoch": 0.21202584, "global_step/max_steps": "640/3019", "percentage": "21.20%", "elapsed_time": "2h 24m 32s", "remaining_time": "8h 57m 17s", "memory(GiB)": 39.75, "train_speed(iter/s)": 0.073795} +{"loss": 0.20494385, "grad_norm": 0.60560387, "learning_rate": 9.14e-06, "tIoU": 0.43162751, "tRecall_50": 0.44375, "tRecall_75": 0.2875, "tRecall_90": 0.18125, "token_acc": 0.92524644, "epoch": 0.21368229, "global_step/max_steps": "645/3019", "percentage": "21.36%", "elapsed_time": "2h 25m 43s", "remaining_time": "8h 56m 21s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073769} +{"loss": 0.18748779, "grad_norm": 0.84822381, "learning_rate": 9.13e-06, "tIoU": 0.46530313, "tRecall_50": 0.53125, "tRecall_75": 0.3, "tRecall_90": 0.175, "token_acc": 0.93335154, "epoch": 0.21533874, "global_step/max_steps": "650/3019", "percentage": "21.53%", "elapsed_time": "2h 26m 54s", "remaining_time": "8h 55m 25s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073743} +{"loss": 0.19372559, "grad_norm": 0.54338378, "learning_rate": 9.11e-06, "tIoU": 0.49716048, "tRecall_50": 0.5625, "tRecall_75": 0.375, "tRecall_90": 0.19375, "token_acc": 0.93183662, "epoch": 0.2169952, "global_step/max_steps": "655/3019", "percentage": "21.70%", "elapsed_time": "2h 28m 0s", "remaining_time": "8h 54m 9s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073761} +{"loss": 0.19257812, "grad_norm": 0.68062013, "learning_rate": 9.1e-06, "tIoU": 0.45461712, "tRecall_50": 0.4875, "tRecall_75": 0.30625, "tRecall_90": 0.1625, "token_acc": 0.92556371, "epoch": 0.21865165, "global_step/max_steps": "660/3019", "percentage": "21.86%", "elapsed_time": "2h 29m 8s", "remaining_time": "8h 53m 3s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073758} +{"loss": 0.19219971, "grad_norm": 0.58294481, "learning_rate": 9.08e-06, "tIoU": 0.45134282, "tRecall_50": 0.48125, "tRecall_75": 0.29375, "tRecall_90": 0.175, "token_acc": 0.92893952, "epoch": 0.2203081, "global_step/max_steps": "665/3019", "percentage": "22.03%", "elapsed_time": "2h 30m 16s", "remaining_time": "8h 51m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073754} +{"loss": 0.20467529, "grad_norm": 0.86679602, "learning_rate": 9.07e-06, "tIoU": 0.42314701, "tRecall_50": 0.45, "tRecall_75": 0.2625, "tRecall_90": 0.15625, "token_acc": 0.92488391, "epoch": 0.22196455, "global_step/max_steps": "670/3019", "percentage": "22.19%", "elapsed_time": "2h 31m 22s", "remaining_time": "8h 50m 43s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073767} +{"loss": 0.19257812, "grad_norm": 0.72036397, "learning_rate": 9.05e-06, "tIoU": 0.46733718, "tRecall_50": 0.4625, "tRecall_75": 0.33125, "tRecall_90": 0.2375, "token_acc": 0.93151432, "epoch": 0.223621, "global_step/max_steps": "675/3019", "percentage": "22.36%", "elapsed_time": "2h 32m 29s", "remaining_time": "8h 49m 32s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073775} +{"loss": 0.18389893, "grad_norm": 0.6191057, "learning_rate": 9.03e-06, "tIoU": 0.48230271, "tRecall_50": 0.50625, "tRecall_75": 0.34375, "tRecall_90": 0.23125, "token_acc": 0.9328951, "epoch": 0.22527746, "global_step/max_steps": "680/3019", "percentage": "22.52%", "elapsed_time": "2h 33m 36s", "remaining_time": "8h 48m 22s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073779} +{"loss": 0.18983154, "grad_norm": 1.02312183, "learning_rate": 9.02e-06, "tIoU": 0.47816577, "tRecall_50": 0.5375, "tRecall_75": 0.31875, "tRecall_90": 0.16875, "token_acc": 0.92833787, "epoch": 0.22693391, "global_step/max_steps": "685/3019", "percentage": "22.69%", "elapsed_time": "2h 34m 43s", "remaining_time": "8h 47m 10s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07379} +{"loss": 0.19080811, "grad_norm": 0.60806262, "learning_rate": 9e-06, "tIoU": 0.42686448, "tRecall_50": 0.45, "tRecall_75": 0.2875, "tRecall_90": 0.175, "token_acc": 0.93041096, "epoch": 0.22859036, "global_step/max_steps": "690/3019", "percentage": "22.86%", "elapsed_time": "2h 35m 47s", "remaining_time": "8h 45m 51s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073816} +{"loss": 0.18778076, "grad_norm": 0.63282341, "learning_rate": 8.99e-06, "tIoU": 0.46008978, "tRecall_50": 0.48125, "tRecall_75": 0.36875, "tRecall_90": 0.20625, "token_acc": 0.932208, "epoch": 0.23024681, "global_step/max_steps": "695/3019", "percentage": "23.02%", "elapsed_time": "2h 36m 53s", "remaining_time": "8h 44m 38s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073828} +{"loss": 0.196521, "grad_norm": 0.58549178, "learning_rate": 8.97e-06, "tIoU": 0.44810834, "tRecall_50": 0.48125, "tRecall_75": 0.3125, "tRecall_90": 0.1875, "token_acc": 0.9253078, "epoch": 0.23190326, "global_step/max_steps": "700/3019", "percentage": "23.19%", "elapsed_time": "2h 38m 1s", "remaining_time": "8h 43m 32s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073825} +{"loss": 0.18496704, "grad_norm": 0.5636034, "learning_rate": 8.95e-06, "tIoU": 0.48276019, "tRecall_50": 0.5, "tRecall_75": 0.3, "tRecall_90": 0.175, "token_acc": 0.92915755, "epoch": 0.23355972, "global_step/max_steps": "705/3019", "percentage": "23.35%", "elapsed_time": "2h 39m 8s", "remaining_time": "8h 42m 20s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073835} +{"loss": 0.17851563, "grad_norm": 0.6367529, "learning_rate": 8.94e-06, "tIoU": 0.46496215, "tRecall_50": 0.5375, "tRecall_75": 0.3, "tRecall_90": 0.19375, "token_acc": 0.93053664, "epoch": 0.23521617, "global_step/max_steps": "710/3019", "percentage": "23.52%", "elapsed_time": "2h 40m 13s", "remaining_time": "8h 41m 3s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073856} +{"loss": 0.17714844, "grad_norm": 0.53599274, "learning_rate": 8.92e-06, "tIoU": 0.50051703, "tRecall_50": 0.55, "tRecall_75": 0.33125, "tRecall_90": 0.20625, "token_acc": 0.93446204, "epoch": 0.23687262, "global_step/max_steps": "715/3019", "percentage": "23.68%", "elapsed_time": "2h 41m 22s", "remaining_time": "8h 40m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073843} +{"loss": 0.18304443, "grad_norm": 1.33014357, "learning_rate": 8.9e-06, "tIoU": 0.47890749, "tRecall_50": 0.525, "tRecall_75": 0.3375, "tRecall_90": 0.175, "token_acc": 0.93309667, "epoch": 0.23852907, "global_step/max_steps": "720/3019", "percentage": "23.85%", "elapsed_time": "2h 42m 30s", "remaining_time": "8h 38m 52s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073845} +{"loss": 0.19580078, "grad_norm": 0.98491228, "learning_rate": 8.89e-06, "tIoU": 0.44307227, "tRecall_50": 0.4625, "tRecall_75": 0.29375, "tRecall_90": 0.20625, "token_acc": 0.92789116, "epoch": 0.24018552, "global_step/max_steps": "725/3019", "percentage": "24.01%", "elapsed_time": "2h 43m 41s", "remaining_time": "8h 37m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073817} +{"loss": 0.17957764, "grad_norm": 0.43194753, "learning_rate": 8.87e-06, "tIoU": 0.4670526, "tRecall_50": 0.5125, "tRecall_75": 0.3125, "tRecall_90": 0.19375, "token_acc": 0.93313319, "epoch": 0.24184197, "global_step/max_steps": "730/3019", "percentage": "24.18%", "elapsed_time": "2h 44m 51s", "remaining_time": "8h 36m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073799} +{"loss": 0.19373779, "grad_norm": 0.72419053, "learning_rate": 8.85e-06, "tIoU": 0.46418605, "tRecall_50": 0.4875, "tRecall_75": 0.29375, "tRecall_90": 0.19375, "token_acc": 0.93014406, "epoch": 0.24349843, "global_step/max_steps": "735/3019", "percentage": "24.35%", "elapsed_time": "2h 45m 58s", "remaining_time": "8h 35m 44s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073809} +{"loss": 0.18074951, "grad_norm": 1.0203234, "learning_rate": 8.84e-06, "tIoU": 0.47715788, "tRecall_50": 0.50625, "tRecall_75": 0.38125, "tRecall_90": 0.2625, "token_acc": 0.93434207, "epoch": 0.24515488, "global_step/max_steps": "740/3019", "percentage": "24.51%", "elapsed_time": "2h 47m 7s", "remaining_time": "8h 34m 42s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073796} +{"loss": 0.19556885, "grad_norm": 0.66058117, "learning_rate": 8.82e-06, "tIoU": 0.4769938, "tRecall_50": 0.51875, "tRecall_75": 0.3375, "tRecall_90": 0.2125, "token_acc": 0.93241042, "epoch": 0.24681133, "global_step/max_steps": "745/3019", "percentage": "24.68%", "elapsed_time": "2h 48m 20s", "remaining_time": "8h 33m 49s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07376} +{"loss": 0.17875977, "grad_norm": 0.73717785, "learning_rate": 8.8e-06, "tIoU": 0.47754402, "tRecall_50": 0.4875, "tRecall_75": 0.3375, "tRecall_90": 0.2125, "token_acc": 0.93520283, "epoch": 0.24846778, "global_step/max_steps": "750/3019", "percentage": "24.84%", "elapsed_time": "2h 49m 29s", "remaining_time": "8h 32m 46s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07375} +{"loss": 0.18903809, "grad_norm": 1.06428552, "learning_rate": 8.78e-06, "tIoU": 0.48411899, "tRecall_50": 0.53125, "tRecall_75": 0.375, "tRecall_90": 0.20625, "token_acc": 0.93137255, "epoch": 0.25012423, "global_step/max_steps": "755/3019", "percentage": "25.01%", "elapsed_time": "2h 50m 35s", "remaining_time": "8h 31m 32s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073764} +{"loss": 0.17894287, "grad_norm": 0.80853009, "learning_rate": 8.77e-06, "tIoU": 0.47337232, "tRecall_50": 0.4875, "tRecall_75": 0.36875, "tRecall_90": 0.25, "token_acc": 0.93729553, "epoch": 0.25178069, "global_step/max_steps": "760/3019", "percentage": "25.17%", "elapsed_time": "2h 51m 40s", "remaining_time": "8h 30m 16s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073784} +{"loss": 0.19639893, "grad_norm": 1.06046712, "learning_rate": 8.75e-06, "tIoU": 0.46307893, "tRecall_50": 0.49375, "tRecall_75": 0.325, "tRecall_90": 0.175, "token_acc": 0.92863002, "epoch": 0.25343714, "global_step/max_steps": "765/3019", "percentage": "25.34%", "elapsed_time": "2h 52m 45s", "remaining_time": "8h 29m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073802} +{"loss": 0.18775635, "grad_norm": 0.74983865, "learning_rate": 8.73e-06, "tIoU": 0.46113019, "tRecall_50": 0.5125, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.92960043, "epoch": 0.25509359, "global_step/max_steps": "770/3019", "percentage": "25.51%", "elapsed_time": "2h 53m 57s", "remaining_time": "8h 28m 5s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073774} +{"loss": 0.17634277, "grad_norm": 0.71016169, "learning_rate": 8.71e-06, "tIoU": 0.46548405, "tRecall_50": 0.50625, "tRecall_75": 0.34375, "tRecall_90": 0.14375, "token_acc": 0.93093093, "epoch": 0.25675004, "global_step/max_steps": "775/3019", "percentage": "25.67%", "elapsed_time": "2h 55m 2s", "remaining_time": "8h 26m 48s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073795} +{"loss": 0.18895264, "grad_norm": 0.868873, "learning_rate": 8.69e-06, "tIoU": 0.49425182, "tRecall_50": 0.51875, "tRecall_75": 0.3375, "tRecall_90": 0.2, "token_acc": 0.93161465, "epoch": 0.25840649, "global_step/max_steps": "780/3019", "percentage": "25.84%", "elapsed_time": "2h 56m 18s", "remaining_time": "8h 26m 4s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073737} +{"loss": 0.18460693, "grad_norm": 0.45646572, "learning_rate": 8.68e-06, "tIoU": 0.48412971, "tRecall_50": 0.48125, "tRecall_75": 0.38125, "tRecall_90": 0.24375, "token_acc": 0.93367623, "epoch": 0.26006295, "global_step/max_steps": "785/3019", "percentage": "26.00%", "elapsed_time": "2h 57m 23s", "remaining_time": "8h 24m 50s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073752} +{"loss": 0.17938232, "grad_norm": 0.65646106, "learning_rate": 8.66e-06, "tIoU": 0.47668638, "tRecall_50": 0.53125, "tRecall_75": 0.3, "tRecall_90": 0.20625, "token_acc": 0.9309595, "epoch": 0.2617194, "global_step/max_steps": "790/3019", "percentage": "26.17%", "elapsed_time": "2h 58m 34s", "remaining_time": "8h 23m 52s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07373} +{"loss": 0.17736816, "grad_norm": 0.58703083, "learning_rate": 8.64e-06, "tIoU": 0.45902901, "tRecall_50": 0.49375, "tRecall_75": 0.31875, "tRecall_90": 0.21875, "token_acc": 0.93387534, "epoch": 0.26337585, "global_step/max_steps": "795/3019", "percentage": "26.33%", "elapsed_time": "2h 59m 39s", "remaining_time": "8h 22m 35s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07375} +{"loss": 0.19100342, "grad_norm": 0.59435982, "learning_rate": 8.62e-06, "tIoU": 0.44837189, "tRecall_50": 0.48125, "tRecall_75": 0.3375, "tRecall_90": 0.175, "token_acc": 0.92796956, "epoch": 0.2650323, "global_step/max_steps": "800/3019", "percentage": "26.50%", "elapsed_time": "3h 0m 45s", "remaining_time": "8h 21m 22s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073764} +{"loss": 0.19122314, "grad_norm": 1.12884128, "learning_rate": 8.6e-06, "tIoU": 0.44898138, "tRecall_50": 0.48125, "tRecall_75": 0.2875, "tRecall_90": 0.2, "token_acc": 0.93133515, "epoch": 0.26668875, "global_step/max_steps": "805/3019", "percentage": "26.66%", "elapsed_time": "3h 1m 53s", "remaining_time": "8h 20m 16s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073759} +{"loss": 0.18260498, "grad_norm": 0.62520289, "learning_rate": 8.58e-06, "tIoU": 0.48221955, "tRecall_50": 0.5125, "tRecall_75": 0.3125, "tRecall_90": 0.21875, "token_acc": 0.93375136, "epoch": 0.2683452, "global_step/max_steps": "810/3019", "percentage": "26.83%", "elapsed_time": "3h 2m 58s", "remaining_time": "8h 19m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073778} +{"loss": 0.17353516, "grad_norm": 0.63082606, "learning_rate": 8.57e-06, "tIoU": 0.50018773, "tRecall_50": 0.51875, "tRecall_75": 0.38125, "tRecall_90": 0.225, "token_acc": 0.93591492, "epoch": 0.27000166, "global_step/max_steps": "815/3019", "percentage": "27.00%", "elapsed_time": "3h 4m 7s", "remaining_time": "8h 17m 55s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073772} +{"loss": 0.19133301, "grad_norm": 0.96118796, "learning_rate": 8.55e-06, "tIoU": 0.47109642, "tRecall_50": 0.4875, "tRecall_75": 0.375, "tRecall_90": 0.21875, "token_acc": 0.92907801, "epoch": 0.27165811, "global_step/max_steps": "820/3019", "percentage": "27.16%", "elapsed_time": "3h 5m 13s", "remaining_time": "8h 16m 42s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073785} +{"loss": 0.18171387, "grad_norm": 0.62503117, "learning_rate": 8.53e-06, "tIoU": 0.47987251, "tRecall_50": 0.53125, "tRecall_75": 0.36875, "tRecall_90": 0.1875, "token_acc": 0.93468161, "epoch": 0.27331456, "global_step/max_steps": "825/3019", "percentage": "27.33%", "elapsed_time": "3h 6m 21s", "remaining_time": "8h 15m 35s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073785} +{"loss": 0.17475586, "grad_norm": 0.49568808, "learning_rate": 8.51e-06, "tIoU": 0.49635658, "tRecall_50": 0.5375, "tRecall_75": 0.33125, "tRecall_90": 0.21875, "token_acc": 0.93613537, "epoch": 0.27497101, "global_step/max_steps": "830/3019", "percentage": "27.49%", "elapsed_time": "3h 7m 28s", "remaining_time": "8h 14m 25s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073789} +{"loss": 0.18552246, "grad_norm": 1.07614172, "learning_rate": 8.49e-06, "tIoU": 0.44313598, "tRecall_50": 0.46875, "tRecall_75": 0.2875, "tRecall_90": 0.1875, "token_acc": 0.9296276, "epoch": 0.27662746, "global_step/max_steps": "835/3019", "percentage": "27.66%", "elapsed_time": "3h 8m 33s", "remaining_time": "8h 13m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073806} +{"loss": 0.19713135, "grad_norm": 0.70680147, "learning_rate": 8.47e-06, "tIoU": 0.44738026, "tRecall_50": 0.48125, "tRecall_75": 0.26875, "tRecall_90": 0.14375, "token_acc": 0.92663043, "epoch": 0.27828392, "global_step/max_steps": "840/3019", "percentage": "27.82%", "elapsed_time": "3h 9m 39s", "remaining_time": "8h 11m 59s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073816} +{"loss": 0.18066406, "grad_norm": 0.5590986, "learning_rate": 8.45e-06, "tIoU": 0.46116395, "tRecall_50": 0.51875, "tRecall_75": 0.325, "tRecall_90": 0.175, "token_acc": 0.92992061, "epoch": 0.27994037, "global_step/max_steps": "845/3019", "percentage": "27.99%", "elapsed_time": "3h 10m 45s", "remaining_time": "8h 10m 45s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073831} +{"loss": 0.18641357, "grad_norm": 0.72481734, "learning_rate": 8.43e-06, "tIoU": 0.45128007, "tRecall_50": 0.48125, "tRecall_75": 0.30625, "tRecall_90": 0.18125, "token_acc": 0.928785, "epoch": 0.28159682, "global_step/max_steps": "850/3019", "percentage": "28.16%", "elapsed_time": "3h 11m 48s", "remaining_time": "8h 9m 27s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073858} +{"loss": 0.19487305, "grad_norm": 0.41906714, "learning_rate": 8.41e-06, "tIoU": 0.45475745, "tRecall_50": 0.5, "tRecall_75": 0.28125, "tRecall_90": 0.19375, "token_acc": 0.92932862, "epoch": 0.28325327, "global_step/max_steps": "855/3019", "percentage": "28.32%", "elapsed_time": "3h 12m 57s", "remaining_time": "8h 8m 23s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073849} +{"loss": 0.17907104, "grad_norm": 0.57942897, "learning_rate": 8.39e-06, "tIoU": 0.53110037, "tRecall_50": 0.58125, "tRecall_75": 0.35, "tRecall_90": 0.23125, "token_acc": 0.93569482, "epoch": 0.28490972, "global_step/max_steps": "860/3019", "percentage": "28.49%", "elapsed_time": "3h 14m 3s", "remaining_time": "8h 7m 9s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073864} +{"loss": 0.19573975, "grad_norm": 0.60247272, "learning_rate": 8.37e-06, "tIoU": 0.44206543, "tRecall_50": 0.46875, "tRecall_75": 0.26875, "tRecall_90": 0.175, "token_acc": 0.92966194, "epoch": 0.28656618, "global_step/max_steps": "865/3019", "percentage": "28.65%", "elapsed_time": "3h 15m 12s", "remaining_time": "8h 6m 6s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073853} +{"loss": 0.20216064, "grad_norm": 0.88688105, "learning_rate": 8.35e-06, "tIoU": 0.37321641, "tRecall_50": 0.39375, "tRecall_75": 0.25, "tRecall_90": 0.15, "token_acc": 0.92702332, "epoch": 0.28822263, "global_step/max_steps": "870/3019", "percentage": "28.82%", "elapsed_time": "3h 16m 18s", "remaining_time": "8h 4m 55s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07386} +{"loss": 0.18554688, "grad_norm": 0.58249789, "learning_rate": 8.33e-06, "tIoU": 0.42017298, "tRecall_50": 0.45625, "tRecall_75": 0.2875, "tRecall_90": 0.15, "token_acc": 0.93073357, "epoch": 0.28987908, "global_step/max_steps": "875/3019", "percentage": "28.98%", "elapsed_time": "3h 17m 21s", "remaining_time": "8h 3m 36s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07389} +{"loss": 0.18995361, "grad_norm": 0.57931828, "learning_rate": 8.31e-06, "tIoU": 0.48091965, "tRecall_50": 0.49375, "tRecall_75": 0.35625, "tRecall_90": 0.25625, "token_acc": 0.93090316, "epoch": 0.29153553, "global_step/max_steps": "880/3019", "percentage": "29.15%", "elapsed_time": "3h 18m 34s", "remaining_time": "8h 2m 40s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073859} +{"loss": 0.17595215, "grad_norm": 0.44487688, "learning_rate": 8.29e-06, "tIoU": 0.50846953, "tRecall_50": 0.54375, "tRecall_75": 0.325, "tRecall_90": 0.20625, "token_acc": 0.93267381, "epoch": 0.29319198, "global_step/max_steps": "885/3019", "percentage": "29.31%", "elapsed_time": "3h 19m 38s", "remaining_time": "8h 1m 23s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073883} +{"loss": 0.17590332, "grad_norm": 1.13230801, "learning_rate": 8.27e-06, "tIoU": 0.44657159, "tRecall_50": 0.475, "tRecall_75": 0.3, "tRecall_90": 0.175, "token_acc": 0.93132465, "epoch": 0.29484843, "global_step/max_steps": "890/3019", "percentage": "29.48%", "elapsed_time": "3h 20m 44s", "remaining_time": "8h 0m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073894} +{"loss": 0.17841797, "grad_norm": 0.56066835, "learning_rate": 8.25e-06, "tIoU": 0.45934696, "tRecall_50": 0.50625, "tRecall_75": 0.3, "tRecall_90": 0.20625, "token_acc": 0.93555375, "epoch": 0.29650489, "global_step/max_steps": "895/3019", "percentage": "29.65%", "elapsed_time": "3h 21m 47s", "remaining_time": "7h 58m 54s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073919} +{"loss": 0.19599609, "grad_norm": 0.92724705, "learning_rate": 8.23e-06, "tIoU": 0.39311662, "tRecall_50": 0.4125, "tRecall_75": 0.2375, "tRecall_90": 0.14375, "token_acc": 0.92353741, "epoch": 0.29816134, "global_step/max_steps": "900/3019", "percentage": "29.81%", "elapsed_time": "3h 22m 52s", "remaining_time": "7h 57m 38s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07394} +{"loss": 0.18623047, "grad_norm": 0.58631378, "learning_rate": 8.21e-06, "tIoU": 0.46013169, "tRecall_50": 0.475, "tRecall_75": 0.30625, "tRecall_90": 0.19375, "token_acc": 0.92992061, "epoch": 0.29981779, "global_step/max_steps": "905/3019", "percentage": "29.98%", "elapsed_time": "3h 24m 0s", "remaining_time": "7h 56m 32s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073936} +{"loss": 0.19263916, "grad_norm": 0.78811568, "learning_rate": 8.19e-06, "tIoU": 0.43993134, "tRecall_50": 0.5125, "tRecall_75": 0.3, "tRecall_90": 0.15625, "token_acc": 0.92821918, "epoch": 0.30147424, "global_step/max_steps": "910/3019", "percentage": "30.14%", "elapsed_time": "3h 25m 4s", "remaining_time": "7h 55m 15s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073959} +{"loss": 0.17945557, "grad_norm": 0.66208017, "learning_rate": 8.17e-06, "tIoU": 0.50310025, "tRecall_50": 0.55, "tRecall_75": 0.34375, "tRecall_90": 0.20625, "token_acc": 0.93061224, "epoch": 0.30313069, "global_step/max_steps": "915/3019", "percentage": "30.31%", "elapsed_time": "3h 26m 10s", "remaining_time": "7h 54m 5s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073966} +{"loss": 0.1699707, "grad_norm": 0.50188398, "learning_rate": 8.15e-06, "tIoU": 0.52973237, "tRecall_50": 0.58125, "tRecall_75": 0.3875, "tRecall_90": 0.225, "token_acc": 0.9386252, "epoch": 0.30478715, "global_step/max_steps": "920/3019", "percentage": "30.47%", "elapsed_time": "3h 27m 21s", "remaining_time": "7h 53m 4s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073948} +{"loss": 0.19112549, "grad_norm": 0.72237772, "learning_rate": 8.13e-06, "tIoU": 0.49451518, "tRecall_50": 0.525, "tRecall_75": 0.35, "tRecall_90": 0.225, "token_acc": 0.93204675, "epoch": 0.3064436, "global_step/max_steps": "925/3019", "percentage": "30.64%", "elapsed_time": "3h 28m 26s", "remaining_time": "7h 51m 51s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073963} +{"loss": 0.18469849, "grad_norm": 0.57583016, "learning_rate": 8.11e-06, "tIoU": 0.4644866, "tRecall_50": 0.5, "tRecall_75": 0.325, "tRecall_90": 0.2, "token_acc": 0.93196721, "epoch": 0.30810005, "global_step/max_steps": "930/3019", "percentage": "30.80%", "elapsed_time": "3h 29m 31s", "remaining_time": "7h 50m 38s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073977} +{"loss": 0.18205566, "grad_norm": 0.52483892, "learning_rate": 8.09e-06, "tIoU": 0.44404759, "tRecall_50": 0.45625, "tRecall_75": 0.3125, "tRecall_90": 0.1625, "token_acc": 0.92923244, "epoch": 0.3097565, "global_step/max_steps": "935/3019", "percentage": "30.97%", "elapsed_time": "3h 30m 43s", "remaining_time": "7h 49m 40s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073953} +{"loss": 0.19499512, "grad_norm": 0.56302994, "learning_rate": 8.07e-06, "tIoU": 0.47184548, "tRecall_50": 0.4875, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.92785189, "epoch": 0.31141295, "global_step/max_steps": "940/3019", "percentage": "31.14%", "elapsed_time": "3h 31m 48s", "remaining_time": "7h 48m 27s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073966} +{"loss": 0.19881592, "grad_norm": 0.65386003, "learning_rate": 8.04e-06, "tIoU": 0.4347651, "tRecall_50": 0.48125, "tRecall_75": 0.3, "tRecall_90": 0.1625, "token_acc": 0.92812415, "epoch": 0.31306941, "global_step/max_steps": "945/3019", "percentage": "31.30%", "elapsed_time": "3h 32m 52s", "remaining_time": "7h 47m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073987} +{"loss": 0.20108643, "grad_norm": 0.47911352, "learning_rate": 8.02e-06, "tIoU": 0.4366662, "tRecall_50": 0.4375, "tRecall_75": 0.28125, "tRecall_90": 0.18125, "token_acc": 0.92615804, "epoch": 0.31472586, "global_step/max_steps": "950/3019", "percentage": "31.47%", "elapsed_time": "3h 33m 58s", "remaining_time": "7h 46m 0s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073998} +{"loss": 0.19428711, "grad_norm": 0.88248521, "learning_rate": 8e-06, "tIoU": 0.4746552, "tRecall_50": 0.5, "tRecall_75": 0.35, "tRecall_90": 0.19375, "token_acc": 0.92847393, "epoch": 0.31638231, "global_step/max_steps": "955/3019", "percentage": "31.63%", "elapsed_time": "3h 35m 1s", "remaining_time": "7h 44m 44s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07402} +{"loss": 0.17499695, "grad_norm": 1.31644309, "learning_rate": 7.98e-06, "tIoU": 0.45885448, "tRecall_50": 0.46875, "tRecall_75": 0.34375, "tRecall_90": 0.2, "token_acc": 0.93006034, "epoch": 0.31803876, "global_step/max_steps": "960/3019", "percentage": "31.80%", "elapsed_time": "3h 36m 7s", "remaining_time": "7h 43m 33s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074029} +{"loss": 0.18673096, "grad_norm": 0.74464881, "learning_rate": 7.96e-06, "tIoU": 0.43875232, "tRecall_50": 0.45, "tRecall_75": 0.275, "tRecall_90": 0.175, "token_acc": 0.93199129, "epoch": 0.31969521, "global_step/max_steps": "965/3019", "percentage": "31.96%", "elapsed_time": "3h 37m 13s", "remaining_time": "7h 42m 22s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074038} +{"loss": 0.18834229, "grad_norm": 0.40595162, "learning_rate": 7.94e-06, "tIoU": 0.46003795, "tRecall_50": 0.48125, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.93127897, "epoch": 0.32135166, "global_step/max_steps": "970/3019", "percentage": "32.13%", "elapsed_time": "3h 38m 25s", "remaining_time": "7h 41m 23s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074015} +{"loss": 0.16092529, "grad_norm": 0.58294493, "learning_rate": 7.91e-06, "tIoU": 0.51879516, "tRecall_50": 0.575, "tRecall_75": 0.35625, "tRecall_90": 0.2375, "token_acc": 0.94014758, "epoch": 0.32300812, "global_step/max_steps": "975/3019", "percentage": "32.30%", "elapsed_time": "3h 39m 31s", "remaining_time": "7h 40m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074026} +{"loss": 0.18634033, "grad_norm": 0.53897256, "learning_rate": 7.89e-06, "tIoU": 0.45755715, "tRecall_50": 0.45625, "tRecall_75": 0.31875, "tRecall_90": 0.175, "token_acc": 0.93232269, "epoch": 0.32466457, "global_step/max_steps": "980/3019", "percentage": "32.46%", "elapsed_time": "3h 40m 36s", "remaining_time": "7h 39m 0s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074037} +{"loss": 0.17401123, "grad_norm": 0.48650983, "learning_rate": 7.87e-06, "tIoU": 0.51093817, "tRecall_50": 0.5625, "tRecall_75": 0.3625, "tRecall_90": 0.2375, "token_acc": 0.9346049, "epoch": 0.32632102, "global_step/max_steps": "985/3019", "percentage": "32.63%", "elapsed_time": "3h 41m 46s", "remaining_time": "7h 37m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074026} +{"loss": 0.16990967, "grad_norm": 1.08315063, "learning_rate": 7.85e-06, "tIoU": 0.48476033, "tRecall_50": 0.5, "tRecall_75": 0.36875, "tRecall_90": 0.21875, "token_acc": 0.93628657, "epoch": 0.32797747, "global_step/max_steps": "990/3019", "percentage": "32.79%", "elapsed_time": "3h 42m 54s", "remaining_time": "7h 36m 51s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074021} +{"loss": 0.18963623, "grad_norm": 0.87865633, "learning_rate": 7.83e-06, "tIoU": 0.46937437, "tRecall_50": 0.5125, "tRecall_75": 0.3875, "tRecall_90": 0.21875, "token_acc": 0.93160314, "epoch": 0.32963392, "global_step/max_steps": "995/3019", "percentage": "32.96%", "elapsed_time": "3h 44m 10s", "remaining_time": "7h 36m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073974} +{"loss": 0.19133301, "grad_norm": 0.53060782, "learning_rate": 7.8e-06, "tIoU": 0.41149063, "tRecall_50": 0.44375, "tRecall_75": 0.21875, "tRecall_90": 0.13125, "token_acc": 0.92730738, "epoch": 0.33129038, "global_step/max_steps": "1000/3019", "percentage": "33.12%", "elapsed_time": "3h 45m 19s", "remaining_time": "7h 34m 55s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073967} +{"loss": 0.19432373, "grad_norm": 0.56666362, "learning_rate": 7.78e-06, "tIoU": 0.47617607, "tRecall_50": 0.51875, "tRecall_75": 0.3375, "tRecall_90": 0.19375, "token_acc": 0.92212006, "epoch": 0.33294683, "global_step/max_steps": "1005/3019", "percentage": "33.29%", "elapsed_time": "3h 46m 25s", "remaining_time": "7h 33m 45s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073975} +{"loss": 0.17784424, "grad_norm": 0.61087978, "learning_rate": 7.76e-06, "tIoU": 0.4799778, "tRecall_50": 0.5, "tRecall_75": 0.35625, "tRecall_90": 0.2125, "token_acc": 0.93160763, "epoch": 0.33460328, "global_step/max_steps": "1010/3019", "percentage": "33.45%", "elapsed_time": "3h 47m 36s", "remaining_time": "7h 32m 43s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07396} +{"loss": 0.18638916, "grad_norm": 0.54095513, "learning_rate": 7.74e-06, "tIoU": 0.44047651, "tRecall_50": 0.48125, "tRecall_75": 0.26875, "tRecall_90": 0.125, "token_acc": 0.92802408, "epoch": 0.33625973, "global_step/max_steps": "1015/3019", "percentage": "33.62%", "elapsed_time": "3h 48m 40s", "remaining_time": "7h 31m 30s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073975} +{"loss": 0.18879395, "grad_norm": 0.51346695, "learning_rate": 7.72e-06, "tIoU": 0.4650414, "tRecall_50": 0.5, "tRecall_75": 0.3, "tRecall_90": 0.1375, "token_acc": 0.93189867, "epoch": 0.33791618, "global_step/max_steps": "1020/3019", "percentage": "33.79%", "elapsed_time": "3h 49m 47s", "remaining_time": "7h 30m 21s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073978} +{"loss": 0.18924561, "grad_norm": 1.02830923, "learning_rate": 7.69e-06, "tIoU": 0.44439902, "tRecall_50": 0.4625, "tRecall_75": 0.28125, "tRecall_90": 0.20625, "token_acc": 0.93038493, "epoch": 0.33957264, "global_step/max_steps": "1025/3019", "percentage": "33.95%", "elapsed_time": "3h 50m 56s", "remaining_time": "7h 29m 15s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073973} +{"loss": 0.17459717, "grad_norm": 0.55050838, "learning_rate": 7.67e-06, "tIoU": 0.50164127, "tRecall_50": 0.50625, "tRecall_75": 0.425, "tRecall_90": 0.275, "token_acc": 0.93532609, "epoch": 0.34122909, "global_step/max_steps": "1030/3019", "percentage": "34.12%", "elapsed_time": "3h 52m 3s", "remaining_time": "7h 28m 7s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073974} +{"loss": 0.17758789, "grad_norm": 0.69128388, "learning_rate": 7.65e-06, "tIoU": 0.48855109, "tRecall_50": 0.56875, "tRecall_75": 0.34375, "tRecall_90": 0.21875, "token_acc": 0.93568521, "epoch": 0.34288554, "global_step/max_steps": "1035/3019", "percentage": "34.28%", "elapsed_time": "3h 53m 12s", "remaining_time": "7h 27m 3s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073966} +{"loss": 0.17293701, "grad_norm": 0.77806026, "learning_rate": 7.62e-06, "tIoU": 0.49210291, "tRecall_50": 0.50625, "tRecall_75": 0.375, "tRecall_90": 0.23125, "token_acc": 0.93654267, "epoch": 0.34454199, "global_step/max_steps": "1040/3019", "percentage": "34.45%", "elapsed_time": "3h 54m 20s", "remaining_time": "7h 25m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073964} +{"loss": 0.17843628, "grad_norm": 0.98760092, "learning_rate": 7.6e-06, "tIoU": 0.53149781, "tRecall_50": 0.575, "tRecall_75": 0.3875, "tRecall_90": 0.2375, "token_acc": 0.93816399, "epoch": 0.34619844, "global_step/max_steps": "1045/3019", "percentage": "34.61%", "elapsed_time": "3h 55m 27s", "remaining_time": "7h 24m 47s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073967} +{"loss": 0.18966064, "grad_norm": 0.82028234, "learning_rate": 7.58e-06, "tIoU": 0.49765325, "tRecall_50": 0.5375, "tRecall_75": 0.325, "tRecall_90": 0.175, "token_acc": 0.93360768, "epoch": 0.34785489, "global_step/max_steps": "1050/3019", "percentage": "34.78%", "elapsed_time": "3h 56m 33s", "remaining_time": "7h 23m 36s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073976} +{"loss": 0.19049072, "grad_norm": 0.70272309, "learning_rate": 7.56e-06, "tIoU": 0.47070408, "tRecall_50": 0.5, "tRecall_75": 0.31875, "tRecall_90": 0.16875, "token_acc": 0.92839641, "epoch": 0.34951135, "global_step/max_steps": "1055/3019", "percentage": "34.95%", "elapsed_time": "3h 57m 42s", "remaining_time": "7h 22m 31s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073971} +{"loss": 0.18425293, "grad_norm": 0.59523332, "learning_rate": 7.53e-06, "tIoU": 0.50456815, "tRecall_50": 0.55, "tRecall_75": 0.3625, "tRecall_90": 0.2125, "token_acc": 0.928785, "epoch": 0.3511678, "global_step/max_steps": "1060/3019", "percentage": "35.11%", "elapsed_time": "3h 58m 54s", "remaining_time": "7h 21m 32s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073945} +{"loss": 0.1784668, "grad_norm": 0.68149555, "learning_rate": 7.51e-06, "tIoU": 0.51221042, "tRecall_50": 0.5875, "tRecall_75": 0.34375, "tRecall_90": 0.19375, "token_acc": 0.93275252, "epoch": 0.35282425, "global_step/max_steps": "1065/3019", "percentage": "35.28%", "elapsed_time": "4h 0m 5s", "remaining_time": "7h 20m 29s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073932} +{"loss": 0.1836792, "grad_norm": 0.75921565, "learning_rate": 7.49e-06, "tIoU": 0.48857489, "tRecall_50": 0.525, "tRecall_75": 0.34375, "tRecall_90": 0.23125, "token_acc": 0.93565976, "epoch": 0.3544807, "global_step/max_steps": "1070/3019", "percentage": "35.44%", "elapsed_time": "4h 1m 13s", "remaining_time": "7h 19m 22s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073929} +{"loss": 0.18165283, "grad_norm": 0.69612277, "learning_rate": 7.46e-06, "tIoU": 0.46857433, "tRecall_50": 0.5, "tRecall_75": 0.3125, "tRecall_90": 0.18125, "token_acc": 0.93137255, "epoch": 0.35613715, "global_step/max_steps": "1075/3019", "percentage": "35.61%", "elapsed_time": "4h 2m 18s", "remaining_time": "7h 18m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073942} +{"loss": 0.180896, "grad_norm": 0.55086672, "learning_rate": 7.44e-06, "tIoU": 0.48640642, "tRecall_50": 0.49375, "tRecall_75": 0.36875, "tRecall_90": 0.225, "token_acc": 0.93712411, "epoch": 0.35779361, "global_step/max_steps": "1080/3019", "percentage": "35.77%", "elapsed_time": "4h 3m 28s", "remaining_time": "7h 17m 7s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073931} +{"loss": 0.17441406, "grad_norm": 0.50226408, "learning_rate": 7.42e-06, "tIoU": 0.51869831, "tRecall_50": 0.525, "tRecall_75": 0.375, "tRecall_90": 0.25, "token_acc": 0.93833561, "epoch": 0.35945006, "global_step/max_steps": "1085/3019", "percentage": "35.94%", "elapsed_time": "4h 4m 29s", "remaining_time": "7h 15m 47s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073965} +{"loss": 0.18186035, "grad_norm": 0.59232551, "learning_rate": 7.39e-06, "tIoU": 0.47256951, "tRecall_50": 0.51875, "tRecall_75": 0.33125, "tRecall_90": 0.175, "token_acc": 0.93051771, "epoch": 0.36110651, "global_step/max_steps": "1090/3019", "percentage": "36.10%", "elapsed_time": "4h 5m 36s", "remaining_time": "7h 14m 39s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073967} +{"loss": 0.18186035, "grad_norm": 0.58329296, "learning_rate": 7.37e-06, "tIoU": 0.49889507, "tRecall_50": 0.55625, "tRecall_75": 0.325, "tRecall_90": 0.2, "token_acc": 0.93369863, "epoch": 0.36276296, "global_step/max_steps": "1095/3019", "percentage": "36.27%", "elapsed_time": "4h 6m 43s", "remaining_time": "7h 13m 30s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07397} +{"loss": 0.18322754, "grad_norm": 0.55087638, "learning_rate": 7.35e-06, "tIoU": 0.4765872, "tRecall_50": 0.50625, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.93409586, "epoch": 0.36441941, "global_step/max_steps": "1100/3019", "percentage": "36.44%", "elapsed_time": "4h 7m 55s", "remaining_time": "7h 12m 30s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073948} +{"loss": 0.16252441, "grad_norm": 0.49825698, "learning_rate": 7.32e-06, "tIoU": 0.55129623, "tRecall_50": 0.5875, "tRecall_75": 0.41875, "tRecall_90": 0.26875, "token_acc": 0.94051535, "epoch": 0.36607587, "global_step/max_steps": "1105/3019", "percentage": "36.60%", "elapsed_time": "4h 9m 3s", "remaining_time": "7h 11m 23s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073947} +{"loss": 0.17033691, "grad_norm": 0.49581203, "learning_rate": 7.3e-06, "tIoU": 0.5293457, "tRecall_50": 0.55, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.9389085, "epoch": 0.36773232, "global_step/max_steps": "1110/3019", "percentage": "36.77%", "elapsed_time": "4h 10m 9s", "remaining_time": "7h 10m 13s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073952} +{"loss": 0.17938232, "grad_norm": 0.59630603, "learning_rate": 7.27e-06, "tIoU": 0.51713986, "tRecall_50": 0.575, "tRecall_75": 0.375, "tRecall_90": 0.1875, "token_acc": 0.9303523, "epoch": 0.36938877, "global_step/max_steps": "1115/3019", "percentage": "36.93%", "elapsed_time": "4h 11m 17s", "remaining_time": "7h 9m 6s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073952} +{"loss": 0.19072266, "grad_norm": 0.78403151, "learning_rate": 7.25e-06, "tIoU": 0.45734377, "tRecall_50": 0.46875, "tRecall_75": 0.29375, "tRecall_90": 0.16875, "token_acc": 0.92903752, "epoch": 0.37104522, "global_step/max_steps": "1120/3019", "percentage": "37.10%", "elapsed_time": "4h 12m 25s", "remaining_time": "7h 8m 0s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073948} +{"loss": 0.18948975, "grad_norm": 0.64492822, "learning_rate": 7.23e-06, "tIoU": 0.48002548, "tRecall_50": 0.50625, "tRecall_75": 0.325, "tRecall_90": 0.2, "token_acc": 0.92927095, "epoch": 0.37270167, "global_step/max_steps": "1125/3019", "percentage": "37.26%", "elapsed_time": "4h 13m 30s", "remaining_time": "7h 6m 48s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07396} +{"loss": 0.17431641, "grad_norm": 0.49803525, "learning_rate": 7.2e-06, "tIoU": 0.50759988, "tRecall_50": 0.53125, "tRecall_75": 0.34375, "tRecall_90": 0.25625, "token_acc": 0.9373461, "epoch": 0.37435812, "global_step/max_steps": "1130/3019", "percentage": "37.43%", "elapsed_time": "4h 14m 40s", "remaining_time": "7h 5m 43s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073952} +{"loss": 0.18666382, "grad_norm": 1.39926589, "learning_rate": 7.18e-06, "tIoU": 0.4678297, "tRecall_50": 0.475, "tRecall_75": 0.3375, "tRecall_90": 0.20625, "token_acc": 0.93257514, "epoch": 0.37601458, "global_step/max_steps": "1135/3019", "percentage": "37.60%", "elapsed_time": "4h 15m 48s", "remaining_time": "7h 4m 36s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073951} +{"loss": 0.1911499, "grad_norm": 0.58237725, "learning_rate": 7.15e-06, "tIoU": 0.45382953, "tRecall_50": 0.46875, "tRecall_75": 0.31875, "tRecall_90": 0.20625, "token_acc": 0.92870697, "epoch": 0.37767103, "global_step/max_steps": "1140/3019", "percentage": "37.76%", "elapsed_time": "4h 16m 57s", "remaining_time": "7h 3m 31s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073942} +{"loss": 0.17542725, "grad_norm": 0.55770987, "learning_rate": 7.13e-06, "tIoU": 0.45658197, "tRecall_50": 0.49375, "tRecall_75": 0.325, "tRecall_90": 0.2125, "token_acc": 0.93286219, "epoch": 0.37932748, "global_step/max_steps": "1145/3019", "percentage": "37.93%", "elapsed_time": "4h 18m 8s", "remaining_time": "7h 2m 29s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073927} +{"loss": 0.16816406, "grad_norm": 0.53211665, "learning_rate": 7.1e-06, "tIoU": 0.541117, "tRecall_50": 0.61875, "tRecall_75": 0.35, "tRecall_90": 0.2125, "token_acc": 0.93543092, "epoch": 0.38098393, "global_step/max_steps": "1150/3019", "percentage": "38.09%", "elapsed_time": "4h 19m 21s", "remaining_time": "7h 1m 30s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073901} +{"loss": 0.17304687, "grad_norm": 0.54871589, "learning_rate": 7.08e-06, "tIoU": 0.53755889, "tRecall_50": 0.5875, "tRecall_75": 0.36875, "tRecall_90": 0.25, "token_acc": 0.93785773, "epoch": 0.38264038, "global_step/max_steps": "1155/3019", "percentage": "38.26%", "elapsed_time": "4h 20m 24s", "remaining_time": "7h 0m 16s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073921} +{"loss": 0.18504639, "grad_norm": 0.64088851, "learning_rate": 7.06e-06, "tIoU": 0.44453301, "tRecall_50": 0.4625, "tRecall_75": 0.28125, "tRecall_90": 0.16875, "token_acc": 0.93032787, "epoch": 0.38429684, "global_step/max_steps": "1160/3019", "percentage": "38.42%", "elapsed_time": "4h 21m 33s", "remaining_time": "6h 59m 9s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073917} +{"loss": 0.16911621, "grad_norm": 0.47475654, "learning_rate": 7.03e-06, "tIoU": 0.47047572, "tRecall_50": 0.50625, "tRecall_75": 0.35, "tRecall_90": 0.2, "token_acc": 0.93527332, "epoch": 0.38595329, "global_step/max_steps": "1165/3019", "percentage": "38.59%", "elapsed_time": "4h 22m 40s", "remaining_time": "6h 58m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073919} +{"loss": 0.19022217, "grad_norm": 0.76273847, "learning_rate": 7.01e-06, "tIoU": 0.48237972, "tRecall_50": 0.51875, "tRecall_75": 0.31875, "tRecall_90": 0.2, "token_acc": 0.93028322, "epoch": 0.38760974, "global_step/max_steps": "1170/3019", "percentage": "38.75%", "elapsed_time": "4h 23m 49s", "remaining_time": "6h 56m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073913} +{"loss": 0.17056885, "grad_norm": 0.7071656, "learning_rate": 6.98e-06, "tIoU": 0.50115442, "tRecall_50": 0.55, "tRecall_75": 0.325, "tRecall_90": 0.2, "token_acc": 0.93566775, "epoch": 0.38926619, "global_step/max_steps": "1175/3019", "percentage": "38.92%", "elapsed_time": "4h 25m 4s", "remaining_time": "6h 56m 0s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073877} +{"loss": 0.18210449, "grad_norm": 0.64850122, "learning_rate": 6.96e-06, "tIoU": 0.50378766, "tRecall_50": 0.5625, "tRecall_75": 0.3625, "tRecall_90": 0.23125, "token_acc": 0.9321341, "epoch": 0.39092264, "global_step/max_steps": "1180/3019", "percentage": "39.09%", "elapsed_time": "4h 26m 12s", "remaining_time": "6h 54m 53s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073876} +{"loss": 0.1675293, "grad_norm": 0.64036489, "learning_rate": 6.93e-06, "tIoU": 0.50668406, "tRecall_50": 0.55, "tRecall_75": 0.3875, "tRecall_90": 0.2125, "token_acc": 0.93889191, "epoch": 0.3925791, "global_step/max_steps": "1185/3019", "percentage": "39.25%", "elapsed_time": "4h 27m 17s", "remaining_time": "6h 53m 40s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073891} +{"loss": 0.17390137, "grad_norm": 0.58124804, "learning_rate": 6.91e-06, "tIoU": 0.46291218, "tRecall_50": 0.525, "tRecall_75": 0.33125, "tRecall_90": 0.18125, "token_acc": 0.93384155, "epoch": 0.39423555, "global_step/max_steps": "1190/3019", "percentage": "39.42%", "elapsed_time": "4h 28m 25s", "remaining_time": "6h 52m 33s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07389} +{"loss": 0.17183838, "grad_norm": 0.52698576, "learning_rate": 6.88e-06, "tIoU": 0.48483205, "tRecall_50": 0.55625, "tRecall_75": 0.35, "tRecall_90": 0.2, "token_acc": 0.9363214, "epoch": 0.395892, "global_step/max_steps": "1195/3019", "percentage": "39.58%", "elapsed_time": "4h 29m 35s", "remaining_time": "6h 51m 30s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073876} +{"loss": 0.18718262, "grad_norm": 0.72628516, "learning_rate": 6.86e-06, "tIoU": 0.45236516, "tRecall_50": 0.4875, "tRecall_75": 0.3125, "tRecall_90": 0.2125, "token_acc": 0.92930661, "epoch": 0.39754845, "global_step/max_steps": "1200/3019", "percentage": "39.75%", "elapsed_time": "4h 30m 44s", "remaining_time": "6h 50m 23s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073872} +{"loss": 0.19154053, "grad_norm": 0.56832331, "learning_rate": 6.83e-06, "tIoU": 0.43550262, "tRecall_50": 0.425, "tRecall_75": 0.3125, "tRecall_90": 0.21875, "token_acc": 0.93193575, "epoch": 0.3992049, "global_step/max_steps": "1205/3019", "percentage": "39.91%", "elapsed_time": "4h 31m 51s", "remaining_time": "6h 49m 14s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073876} +{"loss": 0.1774231, "grad_norm": 0.70350873, "learning_rate": 6.81e-06, "tIoU": 0.52546601, "tRecall_50": 0.58125, "tRecall_75": 0.3625, "tRecall_90": 0.2125, "token_acc": 0.93512486, "epoch": 0.40086135, "global_step/max_steps": "1210/3019", "percentage": "40.08%", "elapsed_time": "4h 32m 58s", "remaining_time": "6h 48m 6s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073877} +{"loss": 0.16162109, "grad_norm": 0.53436285, "learning_rate": 6.78e-06, "tIoU": 0.54245176, "tRecall_50": 0.55625, "tRecall_75": 0.425, "tRecall_90": 0.29375, "token_acc": 0.93896458, "epoch": 0.40251781, "global_step/max_steps": "1215/3019", "percentage": "40.25%", "elapsed_time": "4h 34m 7s", "remaining_time": "6h 47m 0s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073871} +{"loss": 0.18601074, "grad_norm": 0.37724429, "learning_rate": 6.76e-06, "tIoU": 0.44948196, "tRecall_50": 0.44375, "tRecall_75": 0.3, "tRecall_90": 0.19375, "token_acc": 0.92892291, "epoch": 0.40417426, "global_step/max_steps": "1220/3019", "percentage": "40.41%", "elapsed_time": "4h 35m 11s", "remaining_time": "6h 45m 48s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073887} +{"loss": 0.17764893, "grad_norm": 0.45955455, "learning_rate": 6.73e-06, "tIoU": 0.52401428, "tRecall_50": 0.59375, "tRecall_75": 0.3625, "tRecall_90": 0.18125, "token_acc": 0.93816399, "epoch": 0.40583071, "global_step/max_steps": "1225/3019", "percentage": "40.58%", "elapsed_time": "4h 36m 17s", "remaining_time": "6h 44m 37s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073895} +{"loss": 0.17590332, "grad_norm": 0.70192122, "learning_rate": 6.71e-06, "tIoU": 0.43599753, "tRecall_50": 0.4625, "tRecall_75": 0.2875, "tRecall_90": 0.14375, "token_acc": 0.92952381, "epoch": 0.40748716, "global_step/max_steps": "1230/3019", "percentage": "40.74%", "elapsed_time": "4h 37m 19s", "remaining_time": "6h 43m 21s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07392} +{"loss": 0.15350952, "grad_norm": 0.53947002, "learning_rate": 6.68e-06, "tIoU": 0.56933079, "tRecall_50": 0.5875, "tRecall_75": 0.43125, "tRecall_90": 0.29375, "token_acc": 0.94546934, "epoch": 0.40914361, "global_step/max_steps": "1235/3019", "percentage": "40.91%", "elapsed_time": "4h 38m 34s", "remaining_time": "6h 42m 24s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073887} +{"loss": 0.18217773, "grad_norm": 0.96919596, "learning_rate": 6.66e-06, "tIoU": 0.45535507, "tRecall_50": 0.48125, "tRecall_75": 0.3, "tRecall_90": 0.18125, "token_acc": 0.93091207, "epoch": 0.41080007, "global_step/max_steps": "1240/3019", "percentage": "41.07%", "elapsed_time": "4h 39m 46s", "remaining_time": "6h 41m 23s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073868} +{"loss": 0.16960449, "grad_norm": 0.91970015, "learning_rate": 6.63e-06, "tIoU": 0.53196902, "tRecall_50": 0.5875, "tRecall_75": 0.4, "tRecall_90": 0.225, "token_acc": 0.9379085, "epoch": 0.41245652, "global_step/max_steps": "1245/3019", "percentage": "41.24%", "elapsed_time": "4h 40m 54s", "remaining_time": "6h 40m 15s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073868} +{"loss": 0.18415527, "grad_norm": 0.78103828, "learning_rate": 6.61e-06, "tIoU": 0.497365, "tRecall_50": 0.5125, "tRecall_75": 0.35, "tRecall_90": 0.2625, "token_acc": 0.93472373, "epoch": 0.41411297, "global_step/max_steps": "1250/3019", "percentage": "41.40%", "elapsed_time": "4h 41m 58s", "remaining_time": "6h 39m 3s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073883} +{"loss": 0.17766113, "grad_norm": 0.58467865, "learning_rate": 6.58e-06, "tIoU": 0.49566703, "tRecall_50": 0.50625, "tRecall_75": 0.36875, "tRecall_90": 0.25, "token_acc": 0.93688525, "epoch": 0.41576942, "global_step/max_steps": "1255/3019", "percentage": "41.57%", "elapsed_time": "4h 43m 6s", "remaining_time": "6h 37m 55s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073883} +{"loss": 0.1741333, "grad_norm": 0.51974797, "learning_rate": 6.56e-06, "tIoU": 0.50854154, "tRecall_50": 0.53125, "tRecall_75": 0.36875, "tRecall_90": 0.20625, "token_acc": 0.93214774, "epoch": 0.41742587, "global_step/max_steps": "1260/3019", "percentage": "41.74%", "elapsed_time": "4h 44m 17s", "remaining_time": "6h 36m 52s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073869} +{"loss": 0.1685791, "grad_norm": 0.48281902, "learning_rate": 6.53e-06, "tIoU": 0.55761852, "tRecall_50": 0.61875, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.9379254, "epoch": 0.41908233, "global_step/max_steps": "1265/3019", "percentage": "41.90%", "elapsed_time": "4h 45m 25s", "remaining_time": "6h 35m 45s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073867} +{"loss": 0.18017578, "grad_norm": 0.53950649, "learning_rate": 6.51e-06, "tIoU": 0.49621215, "tRecall_50": 0.54375, "tRecall_75": 0.325, "tRecall_90": 0.23125, "token_acc": 0.93556281, "epoch": 0.42073878, "global_step/max_steps": "1270/3019", "percentage": "42.07%", "elapsed_time": "4h 46m 33s", "remaining_time": "6h 34m 38s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073864} +{"loss": 0.16606445, "grad_norm": 0.66378528, "learning_rate": 6.48e-06, "tIoU": 0.47538433, "tRecall_50": 0.4875, "tRecall_75": 0.38125, "tRecall_90": 0.23125, "token_acc": 0.93656002, "epoch": 0.42239523, "global_step/max_steps": "1275/3019", "percentage": "42.23%", "elapsed_time": "4h 47m 42s", "remaining_time": "6h 33m 32s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073858} +{"loss": 0.15682373, "grad_norm": 0.57235211, "learning_rate": 6.45e-06, "tIoU": 0.52204924, "tRecall_50": 0.575, "tRecall_75": 0.36875, "tRecall_90": 0.2375, "token_acc": 0.94069418, "epoch": 0.42405168, "global_step/max_steps": "1280/3019", "percentage": "42.40%", "elapsed_time": "4h 48m 56s", "remaining_time": "6h 32m 33s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073833} +{"loss": 0.17590332, "grad_norm": 0.54726493, "learning_rate": 6.43e-06, "tIoU": 0.47854605, "tRecall_50": 0.5, "tRecall_75": 0.3125, "tRecall_90": 0.2375, "token_acc": 0.93772097, "epoch": 0.42570813, "global_step/max_steps": "1285/3019", "percentage": "42.56%", "elapsed_time": "4h 50m 6s", "remaining_time": "6h 31m 28s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073824} +{"loss": 0.18299561, "grad_norm": 0.56415212, "learning_rate": 6.4e-06, "tIoU": 0.44356194, "tRecall_50": 0.4625, "tRecall_75": 0.26875, "tRecall_90": 0.18125, "token_acc": 0.93176213, "epoch": 0.42736459, "global_step/max_steps": "1290/3019", "percentage": "42.73%", "elapsed_time": "4h 51m 14s", "remaining_time": "6h 30m 20s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073823} +{"loss": 0.18452148, "grad_norm": 0.4343439, "learning_rate": 6.38e-06, "tIoU": 0.50016904, "tRecall_50": 0.54375, "tRecall_75": 0.3375, "tRecall_90": 0.20625, "token_acc": 0.93549265, "epoch": 0.42902104, "global_step/max_steps": "1295/3019", "percentage": "42.89%", "elapsed_time": "4h 52m 20s", "remaining_time": "6h 29m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073828} +{"loss": 0.18483887, "grad_norm": 0.63403261, "learning_rate": 6.35e-06, "tIoU": 0.51359468, "tRecall_50": 0.5875, "tRecall_75": 0.3375, "tRecall_90": 0.175, "token_acc": 0.933043, "epoch": 0.43067749, "global_step/max_steps": "1300/3019", "percentage": "43.06%", "elapsed_time": "4h 53m 26s", "remaining_time": "6h 28m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073834} +{"loss": 0.18204346, "grad_norm": 0.62888789, "learning_rate": 6.33e-06, "tIoU": 0.44141412, "tRecall_50": 0.45, "tRecall_75": 0.28125, "tRecall_90": 0.175, "token_acc": 0.93062005, "epoch": 0.43233394, "global_step/max_steps": "1305/3019", "percentage": "43.23%", "elapsed_time": "4h 54m 31s", "remaining_time": "6h 26m 50s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073846} +{"loss": 0.16866455, "grad_norm": 0.73569608, "learning_rate": 6.3e-06, "tIoU": 0.46663098, "tRecall_50": 0.525, "tRecall_75": 0.35, "tRecall_90": 0.23125, "token_acc": 0.93688525, "epoch": 0.43399039, "global_step/max_steps": "1310/3019", "percentage": "43.39%", "elapsed_time": "4h 55m 43s", "remaining_time": "6h 25m 47s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07383} +{"loss": 0.17736816, "grad_norm": 0.60258281, "learning_rate": 6.27e-06, "tIoU": 0.48667374, "tRecall_50": 0.54375, "tRecall_75": 0.31875, "tRecall_90": 0.2, "token_acc": 0.93691717, "epoch": 0.43564684, "global_step/max_steps": "1315/3019", "percentage": "43.56%", "elapsed_time": "4h 56m 45s", "remaining_time": "6h 24m 32s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073853} +{"loss": 0.18007813, "grad_norm": 0.42847338, "learning_rate": 6.25e-06, "tIoU": 0.49059696, "tRecall_50": 0.5125, "tRecall_75": 0.36875, "tRecall_90": 0.23125, "token_acc": 0.93237371, "epoch": 0.4373033, "global_step/max_steps": "1320/3019", "percentage": "43.72%", "elapsed_time": "4h 57m 56s", "remaining_time": "6h 23m 28s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073841} +{"loss": 0.18408203, "grad_norm": 0.58624262, "learning_rate": 6.22e-06, "tIoU": 0.46089087, "tRecall_50": 0.5125, "tRecall_75": 0.3, "tRecall_90": 0.18125, "token_acc": 0.93083652, "epoch": 0.43895975, "global_step/max_steps": "1325/3019", "percentage": "43.89%", "elapsed_time": "4h 59m 3s", "remaining_time": "6h 22m 20s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073842} +{"loss": 0.19121094, "grad_norm": 1.02607977, "learning_rate": 6.2e-06, "tIoU": 0.48933477, "tRecall_50": 0.5125, "tRecall_75": 0.3625, "tRecall_90": 0.18125, "token_acc": 0.93061224, "epoch": 0.4406162, "global_step/max_steps": "1330/3019", "percentage": "44.05%", "elapsed_time": "5h 0m 10s", "remaining_time": "6h 21m 12s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073845} +{"loss": 0.17241211, "grad_norm": 0.62268656, "learning_rate": 6.17e-06, "tIoU": 0.48278437, "tRecall_50": 0.525, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.93408096, "epoch": 0.44227265, "global_step/max_steps": "1335/3019", "percentage": "44.22%", "elapsed_time": "5h 1m 13s", "remaining_time": "6h 19m 58s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073863} +{"loss": 0.16252441, "grad_norm": 1.15091681, "learning_rate": 6.14e-06, "tIoU": 0.506003, "tRecall_50": 0.55625, "tRecall_75": 0.35, "tRecall_90": 0.2125, "token_acc": 0.94014758, "epoch": 0.4439291, "global_step/max_steps": "1340/3019", "percentage": "44.39%", "elapsed_time": "5h 2m 20s", "remaining_time": "6h 18m 49s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07387} +{"loss": 0.17143555, "grad_norm": 0.62787712, "learning_rate": 6.12e-06, "tIoU": 0.47065639, "tRecall_50": 0.48125, "tRecall_75": 0.375, "tRecall_90": 0.21875, "token_acc": 0.93420693, "epoch": 0.44558556, "global_step/max_steps": "1345/3019", "percentage": "44.55%", "elapsed_time": "5h 3m 28s", "remaining_time": "6h 17m 43s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073865} +{"loss": 0.1824707, "grad_norm": 0.43770373, "learning_rate": 6.09e-06, "tIoU": 0.43801227, "tRecall_50": 0.45, "tRecall_75": 0.3375, "tRecall_90": 0.2125, "token_acc": 0.93114754, "epoch": 0.44724201, "global_step/max_steps": "1350/3019", "percentage": "44.72%", "elapsed_time": "5h 4m 38s", "remaining_time": "6h 16m 37s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073858} +{"loss": 0.1802002, "grad_norm": 0.87501204, "learning_rate": 6.06e-06, "tIoU": 0.46314354, "tRecall_50": 0.50625, "tRecall_75": 0.29375, "tRecall_90": 0.20625, "token_acc": 0.93502594, "epoch": 0.44889846, "global_step/max_steps": "1355/3019", "percentage": "44.88%", "elapsed_time": "5h 5m 44s", "remaining_time": "6h 15m 27s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073866} +{"loss": 0.17661133, "grad_norm": 1.33994329, "learning_rate": 6.04e-06, "tIoU": 0.46857519, "tRecall_50": 0.50625, "tRecall_75": 0.31875, "tRecall_90": 0.2, "token_acc": 0.93315143, "epoch": 0.45055491, "global_step/max_steps": "1360/3019", "percentage": "45.05%", "elapsed_time": "5h 7m 2s", "remaining_time": "6h 14m 32s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073823} +{"loss": 0.16986084, "grad_norm": 0.38020825, "learning_rate": 6.01e-06, "tIoU": 0.47571068, "tRecall_50": 0.5, "tRecall_75": 0.35, "tRecall_90": 0.18125, "token_acc": 0.93738089, "epoch": 0.45221136, "global_step/max_steps": "1365/3019", "percentage": "45.21%", "elapsed_time": "5h 8m 6s", "remaining_time": "6h 13m 20s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073836} +{"loss": 0.17435303, "grad_norm": 0.54174763, "learning_rate": 5.99e-06, "tIoU": 0.47983022, "tRecall_50": 0.5, "tRecall_75": 0.30625, "tRecall_90": 0.2, "token_acc": 0.93142857, "epoch": 0.45386782, "global_step/max_steps": "1370/3019", "percentage": "45.38%", "elapsed_time": "5h 9m 11s", "remaining_time": "6h 12m 9s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073849} +{"loss": 0.18690796, "grad_norm": 0.61072117, "learning_rate": 5.96e-06, "tIoU": 0.4659955, "tRecall_50": 0.45, "tRecall_75": 0.3375, "tRecall_90": 0.21875, "token_acc": 0.93021993, "epoch": 0.45552427, "global_step/max_steps": "1375/3019", "percentage": "45.54%", "elapsed_time": "5h 10m 14s", "remaining_time": "6h 10m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073867} +{"loss": 0.16348267, "grad_norm": 1.21845651, "learning_rate": 5.93e-06, "tIoU": 0.52463779, "tRecall_50": 0.5625, "tRecall_75": 0.375, "tRecall_90": 0.23125, "token_acc": 0.93835243, "epoch": 0.45718072, "global_step/max_steps": "1380/3019", "percentage": "45.71%", "elapsed_time": "5h 11m 24s", "remaining_time": "6h 9m 51s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073857} +{"loss": 0.17071533, "grad_norm": 0.46208957, "learning_rate": 5.91e-06, "tIoU": 0.50487113, "tRecall_50": 0.54375, "tRecall_75": 0.3625, "tRecall_90": 0.2125, "token_acc": 0.93636862, "epoch": 0.45883717, "global_step/max_steps": "1385/3019", "percentage": "45.88%", "elapsed_time": "5h 12m 31s", "remaining_time": "6h 8m 42s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073861} +{"loss": 0.18258057, "grad_norm": 0.44012129, "learning_rate": 5.88e-06, "tIoU": 0.49883342, "tRecall_50": 0.5375, "tRecall_75": 0.35, "tRecall_90": 0.19375, "token_acc": 0.93413501, "epoch": 0.46049362, "global_step/max_steps": "1390/3019", "percentage": "46.04%", "elapsed_time": "5h 13m 39s", "remaining_time": "6h 7m 35s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073858} +{"loss": 0.1644165, "grad_norm": 0.69602174, "learning_rate": 5.85e-06, "tIoU": 0.52412629, "tRecall_50": 0.55625, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.93845316, "epoch": 0.46215007, "global_step/max_steps": "1395/3019", "percentage": "46.21%", "elapsed_time": "5h 14m 51s", "remaining_time": "6h 6m 32s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073842} +{"loss": 0.17698975, "grad_norm": 1.08933628, "learning_rate": 5.83e-06, "tIoU": 0.49909668, "tRecall_50": 0.55, "tRecall_75": 0.375, "tRecall_90": 0.2, "token_acc": 0.93182438, "epoch": 0.46380653, "global_step/max_steps": "1400/3019", "percentage": "46.37%", "elapsed_time": "5h 16m 2s", "remaining_time": "6h 5m 28s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07383} +{"loss": 0.16626587, "grad_norm": 0.62673169, "learning_rate": 5.8e-06, "tIoU": 0.53348894, "tRecall_50": 0.55625, "tRecall_75": 0.39375, "tRecall_90": 0.25625, "token_acc": 0.93875885, "epoch": 0.46546298, "global_step/max_steps": "1405/3019", "percentage": "46.54%", "elapsed_time": "5h 17m 10s", "remaining_time": "6h 4m 20s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07383} +{"loss": 0.18515625, "grad_norm": 0.51433259, "learning_rate": 5.77e-06, "tIoU": 0.4607996, "tRecall_50": 0.5125, "tRecall_75": 0.29375, "tRecall_90": 0.175, "token_acc": 0.9309589, "epoch": 0.46711943, "global_step/max_steps": "1410/3019", "percentage": "46.70%", "elapsed_time": "5h 18m 15s", "remaining_time": "6h 3m 10s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073839} +{"loss": 0.16951904, "grad_norm": 0.48131135, "learning_rate": 5.75e-06, "tIoU": 0.50516686, "tRecall_50": 0.56875, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.93710863, "epoch": 0.46877588, "global_step/max_steps": "1415/3019", "percentage": "46.87%", "elapsed_time": "5h 19m 25s", "remaining_time": "6h 2m 5s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07383} +{"loss": 0.16013794, "grad_norm": 0.77696866, "learning_rate": 5.72e-06, "tIoU": 0.5131547, "tRecall_50": 0.5625, "tRecall_75": 0.3875, "tRecall_90": 0.24375, "token_acc": 0.9359392, "epoch": 0.47043233, "global_step/max_steps": "1420/3019", "percentage": "47.04%", "elapsed_time": "5h 20m 32s", "remaining_time": "6h 0m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073834} +{"loss": 0.18056641, "grad_norm": 0.51521385, "learning_rate": 5.7e-06, "tIoU": 0.45453091, "tRecall_50": 0.4875, "tRecall_75": 0.325, "tRecall_90": 0.23125, "token_acc": 0.9355102, "epoch": 0.47208879, "global_step/max_steps": "1425/3019", "percentage": "47.20%", "elapsed_time": "5h 21m 38s", "remaining_time": "5h 59m 47s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073838} +{"loss": 0.19742432, "grad_norm": 0.6680764, "learning_rate": 5.67e-06, "tIoU": 0.41425529, "tRecall_50": 0.43125, "tRecall_75": 0.25, "tRecall_90": 0.1625, "token_acc": 0.92107404, "epoch": 0.47374524, "global_step/max_steps": "1430/3019", "percentage": "47.37%", "elapsed_time": "5h 22m 43s", "remaining_time": "5h 58m 36s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073852} +{"loss": 0.16871338, "grad_norm": 0.45692986, "learning_rate": 5.64e-06, "tIoU": 0.53470616, "tRecall_50": 0.58125, "tRecall_75": 0.4125, "tRecall_90": 0.26875, "token_acc": 0.94240695, "epoch": 0.47540169, "global_step/max_steps": "1435/3019", "percentage": "47.53%", "elapsed_time": "5h 23m 48s", "remaining_time": "5h 57m 25s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073862} +{"loss": 0.15567627, "grad_norm": 0.65884155, "learning_rate": 5.62e-06, "tIoU": 0.52040372, "tRecall_50": 0.5625, "tRecall_75": 0.34375, "tRecall_90": 0.23125, "token_acc": 0.93598475, "epoch": 0.47705814, "global_step/max_steps": "1440/3019", "percentage": "47.70%", "elapsed_time": "5h 24m 55s", "remaining_time": "5h 56m 17s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073862} +{"loss": 0.16971436, "grad_norm": 0.6113463, "learning_rate": 5.59e-06, "tIoU": 0.51417456, "tRecall_50": 0.55625, "tRecall_75": 0.4, "tRecall_90": 0.225, "token_acc": 0.9385917, "epoch": 0.47871459, "global_step/max_steps": "1445/3019", "percentage": "47.86%", "elapsed_time": "5h 26m 1s", "remaining_time": "5h 55m 7s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073869} +{"loss": 0.17086182, "grad_norm": 0.45739472, "learning_rate": 5.56e-06, "tIoU": 0.52604828, "tRecall_50": 0.5875, "tRecall_75": 0.3875, "tRecall_90": 0.2375, "token_acc": 0.93954383, "epoch": 0.48037105, "global_step/max_steps": "1450/3019", "percentage": "48.03%", "elapsed_time": "5h 27m 7s", "remaining_time": "5h 53m 58s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073875} +{"loss": 0.17530518, "grad_norm": 0.51082021, "learning_rate": 5.54e-06, "tIoU": 0.51797457, "tRecall_50": 0.55625, "tRecall_75": 0.36875, "tRecall_90": 0.2375, "token_acc": 0.93940225, "epoch": 0.4820275, "global_step/max_steps": "1455/3019", "percentage": "48.19%", "elapsed_time": "5h 28m 19s", "remaining_time": "5h 52m 55s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07386} +{"loss": 0.17376709, "grad_norm": 0.65149206, "learning_rate": 5.51e-06, "tIoU": 0.50328541, "tRecall_50": 0.55, "tRecall_75": 0.38125, "tRecall_90": 0.2125, "token_acc": 0.93322476, "epoch": 0.48368395, "global_step/max_steps": "1460/3019", "percentage": "48.36%", "elapsed_time": "5h 29m 23s", "remaining_time": "5h 51m 43s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073874} +{"loss": 0.18035889, "grad_norm": 0.51661623, "learning_rate": 5.48e-06, "tIoU": 0.52162151, "tRecall_50": 0.5625, "tRecall_75": 0.375, "tRecall_90": 0.275, "token_acc": 0.93523035, "epoch": 0.4853404, "global_step/max_steps": "1465/3019", "percentage": "48.53%", "elapsed_time": "5h 30m 32s", "remaining_time": "5h 50m 37s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073868} +{"loss": 0.1644165, "grad_norm": 0.46739942, "learning_rate": 5.46e-06, "tIoU": 0.50967717, "tRecall_50": 0.58125, "tRecall_75": 0.34375, "tRecall_90": 0.175, "token_acc": 0.93828509, "epoch": 0.48699685, "global_step/max_steps": "1470/3019", "percentage": "48.69%", "elapsed_time": "5h 31m 40s", "remaining_time": "5h 49m 29s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073869} +{"loss": 0.17561035, "grad_norm": 0.51384664, "learning_rate": 5.43e-06, "tIoU": 0.45171261, "tRecall_50": 0.45625, "tRecall_75": 0.3, "tRecall_90": 0.175, "token_acc": 0.9321942, "epoch": 0.4886533, "global_step/max_steps": "1475/3019", "percentage": "48.86%", "elapsed_time": "5h 32m 45s", "remaining_time": "5h 48m 20s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073876} +{"loss": 0.1727478, "grad_norm": 0.381221, "learning_rate": 5.4e-06, "tIoU": 0.50955915, "tRecall_50": 0.55, "tRecall_75": 0.375, "tRecall_90": 0.2375, "token_acc": 0.9372112, "epoch": 0.49030976, "global_step/max_steps": "1480/3019", "percentage": "49.02%", "elapsed_time": "5h 33m 48s", "remaining_time": "5h 47m 6s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073895} +{"loss": 0.18843384, "grad_norm": 0.60656953, "learning_rate": 5.38e-06, "tIoU": 0.44904017, "tRecall_50": 0.45625, "tRecall_75": 0.3, "tRecall_90": 0.18125, "token_acc": 0.93031453, "epoch": 0.49196621, "global_step/max_steps": "1485/3019", "percentage": "49.19%", "elapsed_time": "5h 34m 57s", "remaining_time": "5h 46m 0s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07389} +{"loss": 0.16478271, "grad_norm": 0.98500723, "learning_rate": 5.35e-06, "tIoU": 0.50625739, "tRecall_50": 0.54375, "tRecall_75": 0.36875, "tRecall_90": 0.24375, "token_acc": 0.93592552, "epoch": 0.49362266, "global_step/max_steps": "1490/3019", "percentage": "49.35%", "elapsed_time": "5h 36m 4s", "remaining_time": "5h 44m 51s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073894} +{"loss": 0.17548828, "grad_norm": 0.73549807, "learning_rate": 5.32e-06, "tIoU": 0.52690697, "tRecall_50": 0.58125, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.93571234, "epoch": 0.49527911, "global_step/max_steps": "1495/3019", "percentage": "49.52%", "elapsed_time": "5h 37m 10s", "remaining_time": "5h 43m 43s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073897} +{"loss": 0.17823486, "grad_norm": 0.46933788, "learning_rate": 5.29e-06, "tIoU": 0.46794238, "tRecall_50": 0.50625, "tRecall_75": 0.30625, "tRecall_90": 0.19375, "token_acc": 0.93369714, "epoch": 0.49693556, "global_step/max_steps": "1500/3019", "percentage": "49.69%", "elapsed_time": "5h 38m 15s", "remaining_time": "5h 42m 32s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073909} +{"loss": 0.16635742, "grad_norm": 0.53710186, "learning_rate": 5.27e-06, "tIoU": 0.49074044, "tRecall_50": 0.5625, "tRecall_75": 0.35625, "tRecall_90": 0.23125, "token_acc": 0.93676751, "epoch": 0.49859202, "global_step/max_steps": "1505/3019", "percentage": "49.85%", "elapsed_time": "5h 39m 27s", "remaining_time": "5h 41m 29s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073892} +{"loss": 0.17530518, "grad_norm": 0.62714416, "learning_rate": 5.24e-06, "tIoU": 0.48148627, "tRecall_50": 0.5375, "tRecall_75": 0.34375, "tRecall_90": 0.2, "token_acc": 0.9345026, "epoch": 0.50024847, "global_step/max_steps": "1510/3019", "percentage": "50.02%", "elapsed_time": "5h 40m 38s", "remaining_time": "5h 40m 25s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073879} +{"loss": 0.18417969, "grad_norm": 0.71342576, "learning_rate": 5.21e-06, "tIoU": 0.50607615, "tRecall_50": 0.5125, "tRecall_75": 0.36875, "tRecall_90": 0.24375, "token_acc": 0.93787466, "epoch": 0.50190492, "global_step/max_steps": "1515/3019", "percentage": "50.18%", "elapsed_time": "5h 41m 44s", "remaining_time": "5h 39m 16s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073885} +{"loss": 0.18895264, "grad_norm": 0.56658494, "learning_rate": 5.19e-06, "tIoU": 0.48246069, "tRecall_50": 0.50625, "tRecall_75": 0.34375, "tRecall_90": 0.19375, "token_acc": 0.93306343, "epoch": 0.50356137, "global_step/max_steps": "1520/3019", "percentage": "50.35%", "elapsed_time": "5h 42m 51s", "remaining_time": "5h 38m 7s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073889} +{"loss": 0.17012939, "grad_norm": 0.74153858, "learning_rate": 5.16e-06, "tIoU": 0.50148034, "tRecall_50": 0.5375, "tRecall_75": 0.38125, "tRecall_90": 0.20625, "token_acc": 0.9352459, "epoch": 0.50521782, "global_step/max_steps": "1525/3019", "percentage": "50.51%", "elapsed_time": "5h 44m 1s", "remaining_time": "5h 37m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073881} +{"loss": 0.1737915, "grad_norm": 0.55641842, "learning_rate": 5.13e-06, "tIoU": 0.46170502, "tRecall_50": 0.48125, "tRecall_75": 0.31875, "tRecall_90": 0.19375, "token_acc": 0.93533424, "epoch": 0.50687428, "global_step/max_steps": "1530/3019", "percentage": "50.68%", "elapsed_time": "5h 45m 10s", "remaining_time": "5h 35m 55s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073875} +{"loss": 0.16895752, "grad_norm": 0.57664865, "learning_rate": 5.11e-06, "tIoU": 0.46112266, "tRecall_50": 0.5375, "tRecall_75": 0.30625, "tRecall_90": 0.1625, "token_acc": 0.93823449, "epoch": 0.50853073, "global_step/max_steps": "1535/3019", "percentage": "50.84%", "elapsed_time": "5h 46m 17s", "remaining_time": "5h 34m 47s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073877} +{"loss": 0.17355957, "grad_norm": 0.48594368, "learning_rate": 5.08e-06, "tIoU": 0.47720838, "tRecall_50": 0.51875, "tRecall_75": 0.33125, "tRecall_90": 0.19375, "token_acc": 0.93529092, "epoch": 0.51018718, "global_step/max_steps": "1540/3019", "percentage": "51.01%", "elapsed_time": "5h 47m 22s", "remaining_time": "5h 33m 37s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073886} +{"loss": 0.17591553, "grad_norm": 0.47792771, "learning_rate": 5.05e-06, "tIoU": 0.52786899, "tRecall_50": 0.60625, "tRecall_75": 0.36875, "tRecall_90": 0.20625, "token_acc": 0.93474715, "epoch": 0.51184363, "global_step/max_steps": "1545/3019", "percentage": "51.18%", "elapsed_time": "5h 48m 33s", "remaining_time": "5h 32m 32s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073874} +{"loss": 0.16191406, "grad_norm": 0.49381638, "learning_rate": 5.03e-06, "tIoU": 0.54531789, "tRecall_50": 0.575, "tRecall_75": 0.4125, "tRecall_90": 0.275, "token_acc": 0.94152842, "epoch": 0.51350008, "global_step/max_steps": "1550/3019", "percentage": "51.34%", "elapsed_time": "5h 49m 42s", "remaining_time": "5h 31m 25s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073871} +{"loss": 0.16890869, "grad_norm": 0.71610993, "learning_rate": 5e-06, "tIoU": 0.51696401, "tRecall_50": 0.54375, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.93939394, "epoch": 0.51515653, "global_step/max_steps": "1555/3019", "percentage": "51.51%", "elapsed_time": "5h 50m 51s", "remaining_time": "5h 30m 19s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073867} +{"loss": 0.16268311, "grad_norm": 0.59652781, "learning_rate": 4.97e-06, "tIoU": 0.56040773, "tRecall_50": 0.63125, "tRecall_75": 0.40625, "tRecall_90": 0.23125, "token_acc": 0.93709328, "epoch": 0.51681299, "global_step/max_steps": "1560/3019", "percentage": "51.67%", "elapsed_time": "5h 52m 3s", "remaining_time": "5h 29m 16s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07385} +{"loss": 0.17548828, "grad_norm": 1.26501083, "learning_rate": 4.95e-06, "tIoU": 0.4656291, "tRecall_50": 0.50625, "tRecall_75": 0.3625, "tRecall_90": 0.19375, "token_acc": 0.93336959, "epoch": 0.51846944, "global_step/max_steps": "1565/3019", "percentage": "51.84%", "elapsed_time": "5h 53m 10s", "remaining_time": "5h 28m 7s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073855} +{"loss": 0.17598877, "grad_norm": 0.45799071, "learning_rate": 4.92e-06, "tIoU": 0.50430493, "tRecall_50": 0.55625, "tRecall_75": 0.375, "tRecall_90": 0.175, "token_acc": 0.93244348, "epoch": 0.52012589, "global_step/max_steps": "1570/3019", "percentage": "52.00%", "elapsed_time": "5h 54m 18s", "remaining_time": "5h 27m 0s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073852} +{"loss": 0.17666016, "grad_norm": 0.66684419, "learning_rate": 4.89e-06, "tIoU": 0.49046345, "tRecall_50": 0.50625, "tRecall_75": 0.3875, "tRecall_90": 0.225, "token_acc": 0.93641304, "epoch": 0.52178234, "global_step/max_steps": "1575/3019", "percentage": "52.17%", "elapsed_time": "5h 55m 25s", "remaining_time": "5h 25m 52s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073854} +{"loss": 0.15522461, "grad_norm": 0.4330036, "learning_rate": 4.87e-06, "tIoU": 0.53823342, "tRecall_50": 0.5875, "tRecall_75": 0.375, "tRecall_90": 0.24375, "token_acc": 0.9438971, "epoch": 0.52343879, "global_step/max_steps": "1580/3019", "percentage": "52.34%", "elapsed_time": "5h 56m 31s", "remaining_time": "5h 24m 42s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073862} +{"loss": 0.17924805, "grad_norm": 0.63250232, "learning_rate": 4.84e-06, "tIoU": 0.46796865, "tRecall_50": 0.51875, "tRecall_75": 0.2875, "tRecall_90": 0.15, "token_acc": 0.93226333, "epoch": 0.52509525, "global_step/max_steps": "1585/3019", "percentage": "52.50%", "elapsed_time": "5h 57m 32s", "remaining_time": "5h 23m 29s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073883} +{"loss": 0.17851563, "grad_norm": 0.49443552, "learning_rate": 4.81e-06, "tIoU": 0.50531917, "tRecall_50": 0.50625, "tRecall_75": 0.38125, "tRecall_90": 0.25625, "token_acc": 0.93671577, "epoch": 0.5267517, "global_step/max_steps": "1590/3019", "percentage": "52.67%", "elapsed_time": "5h 58m 43s", "remaining_time": "5h 22m 24s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073871} +{"loss": 0.17664795, "grad_norm": 0.55530083, "learning_rate": 4.79e-06, "tIoU": 0.49562931, "tRecall_50": 0.55, "tRecall_75": 0.34375, "tRecall_90": 0.1875, "token_acc": 0.93525571, "epoch": 0.52840815, "global_step/max_steps": "1595/3019", "percentage": "52.83%", "elapsed_time": "5h 59m 53s", "remaining_time": "5h 21m 18s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073866} +{"loss": 0.16912842, "grad_norm": 0.82524282, "learning_rate": 4.76e-06, "tIoU": 0.49861188, "tRecall_50": 0.55625, "tRecall_75": 0.325, "tRecall_90": 0.18125, "token_acc": 0.93829845, "epoch": 0.5300646, "global_step/max_steps": "1600/3019", "percentage": "53.00%", "elapsed_time": "6h 0m 57s", "remaining_time": "5h 20m 7s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073876} +{"loss": 0.17553711, "grad_norm": 0.57492918, "learning_rate": 4.73e-06, "tIoU": 0.46056161, "tRecall_50": 0.50625, "tRecall_75": 0.3625, "tRecall_90": 0.18125, "token_acc": 0.93336945, "epoch": 0.53172105, "global_step/max_steps": "1605/3019", "percentage": "53.16%", "elapsed_time": "6h 2m 7s", "remaining_time": "5h 19m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07387} +{"loss": 0.17668457, "grad_norm": 0.56136346, "learning_rate": 4.71e-06, "tIoU": 0.47531672, "tRecall_50": 0.5, "tRecall_75": 0.325, "tRecall_90": 0.18125, "token_acc": 0.93237371, "epoch": 0.53337751, "global_step/max_steps": "1610/3019", "percentage": "53.33%", "elapsed_time": "6h 3m 16s", "remaining_time": "5h 17m 55s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073864} +{"loss": 0.17596436, "grad_norm": 0.52558571, "learning_rate": 4.68e-06, "tIoU": 0.51209197, "tRecall_50": 0.55625, "tRecall_75": 0.3875, "tRecall_90": 0.23125, "token_acc": 0.93836924, "epoch": 0.53503396, "global_step/max_steps": "1615/3019", "percentage": "53.49%", "elapsed_time": "6h 4m 24s", "remaining_time": "5h 16m 48s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073862} +{"loss": 0.16297607, "grad_norm": 0.70991915, "learning_rate": 4.65e-06, "tIoU": 0.52863998, "tRecall_50": 0.58125, "tRecall_75": 0.3625, "tRecall_90": 0.25625, "token_acc": 0.94280241, "epoch": 0.53669041, "global_step/max_steps": "1620/3019", "percentage": "53.66%", "elapsed_time": "6h 5m 32s", "remaining_time": "5h 15m 40s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073862} +{"loss": 0.1635376, "grad_norm": 0.50859112, "learning_rate": 4.62e-06, "tIoU": 0.5484652, "tRecall_50": 0.60625, "tRecall_75": 0.36875, "tRecall_90": 0.2125, "token_acc": 0.93331519, "epoch": 0.53834686, "global_step/max_steps": "1625/3019", "percentage": "53.83%", "elapsed_time": "6h 6m 41s", "remaining_time": "5h 14m 34s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073858} +{"loss": 0.16826172, "grad_norm": 0.71921748, "learning_rate": 4.6e-06, "tIoU": 0.51817427, "tRecall_50": 0.55, "tRecall_75": 0.40625, "tRecall_90": 0.21875, "token_acc": 0.93865364, "epoch": 0.54000331, "global_step/max_steps": "1630/3019", "percentage": "53.99%", "elapsed_time": "6h 7m 48s", "remaining_time": "5h 13m 25s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073862} +{"loss": 0.18339844, "grad_norm": 0.4722282, "learning_rate": 4.57e-06, "tIoU": 0.45728669, "tRecall_50": 0.4875, "tRecall_75": 0.28125, "tRecall_90": 0.175, "token_acc": 0.93012507, "epoch": 0.54165976, "global_step/max_steps": "1635/3019", "percentage": "54.16%", "elapsed_time": "6h 8m 54s", "remaining_time": "5h 12m 16s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073866} +{"loss": 0.1663208, "grad_norm": 0.69046837, "learning_rate": 4.54e-06, "tIoU": 0.52245998, "tRecall_50": 0.56875, "tRecall_75": 0.375, "tRecall_90": 0.2375, "token_acc": 0.93877551, "epoch": 0.54331622, "global_step/max_steps": "1640/3019", "percentage": "54.32%", "elapsed_time": "6h 9m 58s", "remaining_time": "5h 11m 6s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073878} +{"loss": 0.17559814, "grad_norm": 0.50946373, "learning_rate": 4.52e-06, "tIoU": 0.51000624, "tRecall_50": 0.50625, "tRecall_75": 0.36875, "tRecall_90": 0.2375, "token_acc": 0.93309667, "epoch": 0.54497267, "global_step/max_steps": "1645/3019", "percentage": "54.49%", "elapsed_time": "6h 11m 8s", "remaining_time": "5h 9m 59s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073873} +{"loss": 0.16832886, "grad_norm": 0.38745856, "learning_rate": 4.49e-06, "tIoU": 0.49676719, "tRecall_50": 0.50625, "tRecall_75": 0.35, "tRecall_90": 0.2375, "token_acc": 0.93763676, "epoch": 0.54662912, "global_step/max_steps": "1650/3019", "percentage": "54.65%", "elapsed_time": "6h 12m 12s", "remaining_time": "5h 8m 49s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073883} +{"loss": 0.15733032, "grad_norm": 0.66163045, "learning_rate": 4.46e-06, "tIoU": 0.54246111, "tRecall_50": 0.5875, "tRecall_75": 0.3875, "tRecall_90": 0.275, "token_acc": 0.94048913, "epoch": 0.54828557, "global_step/max_steps": "1655/3019", "percentage": "54.82%", "elapsed_time": "6h 13m 18s", "remaining_time": "5h 7m 40s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073888} +{"loss": 0.18017578, "grad_norm": 3.82171822, "learning_rate": 4.44e-06, "tIoU": 0.5014461, "tRecall_50": 0.5125, "tRecall_75": 0.3625, "tRecall_90": 0.25, "token_acc": 0.9336235, "epoch": 0.54994202, "global_step/max_steps": "1660/3019", "percentage": "54.99%", "elapsed_time": "6h 14m 24s", "remaining_time": "5h 6m 31s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073894} +{"loss": 0.16463623, "grad_norm": 0.64614552, "learning_rate": 4.41e-06, "tIoU": 0.49542522, "tRecall_50": 0.54375, "tRecall_75": 0.3375, "tRecall_90": 0.2, "token_acc": 0.93758517, "epoch": 0.55159848, "global_step/max_steps": "1665/3019", "percentage": "55.15%", "elapsed_time": "6h 15m 29s", "remaining_time": "5h 5m 20s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073905} +{"loss": 0.16251221, "grad_norm": 0.38401863, "learning_rate": 4.38e-06, "tIoU": 0.52716956, "tRecall_50": 0.55625, "tRecall_75": 0.3875, "tRecall_90": 0.25625, "token_acc": 0.9403883, "epoch": 0.55325493, "global_step/max_steps": "1670/3019", "percentage": "55.32%", "elapsed_time": "6h 16m 34s", "remaining_time": "5h 4m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07391} +{"loss": 0.15831299, "grad_norm": 0.68385017, "learning_rate": 4.36e-06, "tIoU": 0.54369922, "tRecall_50": 0.5875, "tRecall_75": 0.4375, "tRecall_90": 0.2375, "token_acc": 0.94033935, "epoch": 0.55491138, "global_step/max_steps": "1675/3019", "percentage": "55.48%", "elapsed_time": "6h 17m 43s", "remaining_time": "5h 3m 5s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073907} +{"loss": 0.17099609, "grad_norm": 0.48298609, "learning_rate": 4.33e-06, "tIoU": 0.50997949, "tRecall_50": 0.525, "tRecall_75": 0.4, "tRecall_90": 0.23125, "token_acc": 0.93443961, "epoch": 0.55656783, "global_step/max_steps": "1680/3019", "percentage": "55.65%", "elapsed_time": "6h 18m 56s", "remaining_time": "5h 2m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073891} +{"loss": 0.18450928, "grad_norm": 0.61358565, "learning_rate": 4.3e-06, "tIoU": 0.48418674, "tRecall_50": 0.55, "tRecall_75": 0.3625, "tRecall_90": 0.16875, "token_acc": 0.93146587, "epoch": 0.55822428, "global_step/max_steps": "1685/3019", "percentage": "55.81%", "elapsed_time": "6h 20m 1s", "remaining_time": "5h 0m 51s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073899} +{"loss": 0.16536865, "grad_norm": 0.57208633, "learning_rate": 4.28e-06, "tIoU": 0.54747376, "tRecall_50": 0.6, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.9392779, "epoch": 0.55988074, "global_step/max_steps": "1690/3019", "percentage": "55.98%", "elapsed_time": "6h 21m 5s", "remaining_time": "4h 59m 40s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073912} +{"loss": 0.17178955, "grad_norm": 0.54203796, "learning_rate": 4.25e-06, "tIoU": 0.50926113, "tRecall_50": 0.5625, "tRecall_75": 0.3625, "tRecall_90": 0.24375, "token_acc": 0.93424658, "epoch": 0.56153719, "global_step/max_steps": "1695/3019", "percentage": "56.14%", "elapsed_time": "6h 22m 6s", "remaining_time": "4h 58m 28s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073932} +{"loss": 0.1845459, "grad_norm": 0.63369238, "learning_rate": 4.23e-06, "tIoU": 0.49170051, "tRecall_50": 0.55625, "tRecall_75": 0.3875, "tRecall_90": 0.2125, "token_acc": 0.93617021, "epoch": 0.56319364, "global_step/max_steps": "1700/3019", "percentage": "56.31%", "elapsed_time": "6h 23m 11s", "remaining_time": "4h 57m 18s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073941} +{"loss": 0.18272705, "grad_norm": 0.90513116, "learning_rate": 4.2e-06, "tIoU": 0.5123229, "tRecall_50": 0.58125, "tRecall_75": 0.36875, "tRecall_90": 0.20625, "token_acc": 0.93388203, "epoch": 0.56485009, "global_step/max_steps": "1705/3019", "percentage": "56.48%", "elapsed_time": "6h 24m 17s", "remaining_time": "4h 56m 9s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073945} +{"loss": 0.16046753, "grad_norm": 0.73996127, "learning_rate": 4.17e-06, "tIoU": 0.55894713, "tRecall_50": 0.61875, "tRecall_75": 0.39375, "tRecall_90": 0.25, "token_acc": 0.94455067, "epoch": 0.56650654, "global_step/max_steps": "1710/3019", "percentage": "56.64%", "elapsed_time": "6h 25m 30s", "remaining_time": "4h 55m 6s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073928} +{"loss": 0.18409424, "grad_norm": 0.41440925, "learning_rate": 4.15e-06, "tIoU": 0.46752391, "tRecall_50": 0.525, "tRecall_75": 0.35, "tRecall_90": 0.19375, "token_acc": 0.92822186, "epoch": 0.56816299, "global_step/max_steps": "1715/3019", "percentage": "56.81%", "elapsed_time": "6h 26m 37s", "remaining_time": "4h 53m 57s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073931} +{"loss": 0.16766968, "grad_norm": 0.47274792, "learning_rate": 4.12e-06, "tIoU": 0.52232771, "tRecall_50": 0.5875, "tRecall_75": 0.35, "tRecall_90": 0.19375, "token_acc": 0.93807965, "epoch": 0.56981945, "global_step/max_steps": "1720/3019", "percentage": "56.97%", "elapsed_time": "6h 27m 42s", "remaining_time": "4h 52m 48s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073938} +{"loss": 0.16210938, "grad_norm": 0.49913594, "learning_rate": 4.09e-06, "tIoU": 0.45806708, "tRecall_50": 0.50625, "tRecall_75": 0.325, "tRecall_90": 0.21875, "token_acc": 0.93799293, "epoch": 0.5714759, "global_step/max_steps": "1725/3019", "percentage": "57.14%", "elapsed_time": "6h 28m 44s", "remaining_time": "4h 51m 36s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073956} +{"loss": 0.18353271, "grad_norm": 1.1948216, "learning_rate": 4.07e-06, "tIoU": 0.52888446, "tRecall_50": 0.5625, "tRecall_75": 0.38125, "tRecall_90": 0.25, "token_acc": 0.93843639, "epoch": 0.57313235, "global_step/max_steps": "1730/3019", "percentage": "57.30%", "elapsed_time": "6h 29m 53s", "remaining_time": "4h 50m 30s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073952} +{"loss": 0.15739746, "grad_norm": 0.46462679, "learning_rate": 4.04e-06, "tIoU": 0.55867105, "tRecall_50": 0.6125, "tRecall_75": 0.4125, "tRecall_90": 0.275, "token_acc": 0.9410793, "epoch": 0.5747888, "global_step/max_steps": "1735/3019", "percentage": "57.47%", "elapsed_time": "6h 30m 58s", "remaining_time": "4h 49m 20s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073961} +{"loss": 0.16785889, "grad_norm": 0.5275194, "learning_rate": 4.01e-06, "tIoU": 0.50469651, "tRecall_50": 0.5625, "tRecall_75": 0.38125, "tRecall_90": 0.20625, "token_acc": 0.93826823, "epoch": 0.57644525, "global_step/max_steps": "1740/3019", "percentage": "57.63%", "elapsed_time": "6h 32m 3s", "remaining_time": "4h 48m 10s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073969} +{"loss": 0.15900879, "grad_norm": 0.52004707, "learning_rate": 3.99e-06, "tIoU": 0.54808693, "tRecall_50": 0.63125, "tRecall_75": 0.38125, "tRecall_90": 0.25, "token_acc": 0.94039194, "epoch": 0.57810171, "global_step/max_steps": "1745/3019", "percentage": "57.80%", "elapsed_time": "6h 33m 14s", "remaining_time": "4h 47m 5s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073959} +{"loss": 0.17436523, "grad_norm": 0.57665187, "learning_rate": 3.96e-06, "tIoU": 0.50860624, "tRecall_50": 0.53125, "tRecall_75": 0.3375, "tRecall_90": 0.2375, "token_acc": 0.9362224, "epoch": 0.57975816, "global_step/max_steps": "1750/3019", "percentage": "57.97%", "elapsed_time": "6h 34m 20s", "remaining_time": "4h 45m 57s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073963} +{"loss": 0.15159912, "grad_norm": 0.59857738, "learning_rate": 3.94e-06, "tIoU": 0.57210455, "tRecall_50": 0.65, "tRecall_75": 0.40625, "tRecall_90": 0.2625, "token_acc": 0.94128891, "epoch": 0.58141461, "global_step/max_steps": "1755/3019", "percentage": "58.13%", "elapsed_time": "6h 35m 25s", "remaining_time": "4h 44m 47s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073971} +{"loss": 0.16317139, "grad_norm": 0.98725176, "learning_rate": 3.91e-06, "tIoU": 0.53881125, "tRecall_50": 0.58125, "tRecall_75": 0.425, "tRecall_90": 0.29375, "token_acc": 0.94379582, "epoch": 0.58307106, "global_step/max_steps": "1760/3019", "percentage": "58.30%", "elapsed_time": "6h 36m 35s", "remaining_time": "4h 43m 41s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073963} +{"loss": 0.14564209, "grad_norm": 0.51408499, "learning_rate": 3.88e-06, "tIoU": 0.54856167, "tRecall_50": 0.58125, "tRecall_75": 0.44375, "tRecall_90": 0.28125, "token_acc": 0.94430794, "epoch": 0.58472751, "global_step/max_steps": "1765/3019", "percentage": "58.46%", "elapsed_time": "6h 37m 38s", "remaining_time": "4h 42m 31s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073976} +{"loss": 0.17509766, "grad_norm": 0.62633908, "learning_rate": 3.86e-06, "tIoU": 0.50564933, "tRecall_50": 0.54375, "tRecall_75": 0.4, "tRecall_90": 0.21875, "token_acc": 0.93535188, "epoch": 0.58638397, "global_step/max_steps": "1770/3019", "percentage": "58.63%", "elapsed_time": "6h 38m 48s", "remaining_time": "4h 41m 24s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073972} +{"loss": 0.17532959, "grad_norm": 0.68491572, "learning_rate": 3.83e-06, "tIoU": 0.51503901, "tRecall_50": 0.5875, "tRecall_75": 0.36875, "tRecall_90": 0.2125, "token_acc": 0.93249522, "epoch": 0.58804042, "global_step/max_steps": "1775/3019", "percentage": "58.79%", "elapsed_time": "6h 40m 0s", "remaining_time": "4h 40m 20s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073957} +{"loss": 0.14559326, "grad_norm": 0.84995371, "learning_rate": 3.8e-06, "tIoU": 0.57193255, "tRecall_50": 0.625, "tRecall_75": 0.425, "tRecall_90": 0.325, "token_acc": 0.94574064, "epoch": 0.58969687, "global_step/max_steps": "1780/3019", "percentage": "58.96%", "elapsed_time": "6h 41m 7s", "remaining_time": "4h 39m 12s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073958} +{"loss": 0.16452637, "grad_norm": 0.5121057, "learning_rate": 3.78e-06, "tIoU": 0.51438451, "tRecall_50": 0.5625, "tRecall_75": 0.40625, "tRecall_90": 0.23125, "token_acc": 0.9362224, "epoch": 0.59135332, "global_step/max_steps": "1785/3019", "percentage": "59.13%", "elapsed_time": "6h 42m 13s", "remaining_time": "4h 38m 3s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073963} +{"loss": 0.17528076, "grad_norm": 0.75347906, "learning_rate": 3.75e-06, "tIoU": 0.49319386, "tRecall_50": 0.5375, "tRecall_75": 0.31875, "tRecall_90": 0.16875, "token_acc": 0.93282567, "epoch": 0.59300977, "global_step/max_steps": "1790/3019", "percentage": "59.29%", "elapsed_time": "6h 43m 24s", "remaining_time": "4h 36m 58s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073954} +{"loss": 0.16246338, "grad_norm": 0.5419091, "learning_rate": 3.73e-06, "tIoU": 0.54563079, "tRecall_50": 0.6375, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.93865364, "epoch": 0.59466622, "global_step/max_steps": "1795/3019", "percentage": "59.46%", "elapsed_time": "6h 44m 35s", "remaining_time": "4h 35m 53s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073942} +{"loss": 0.18109131, "grad_norm": 0.55788028, "learning_rate": 3.7e-06, "tIoU": 0.4708672, "tRecall_50": 0.51875, "tRecall_75": 0.29375, "tRecall_90": 0.14375, "token_acc": 0.93161465, "epoch": 0.59632268, "global_step/max_steps": "1800/3019", "percentage": "59.62%", "elapsed_time": "6h 45m 41s", "remaining_time": "4h 34m 44s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073949} +{"loss": 0.16984253, "grad_norm": 1.20913374, "learning_rate": 3.67e-06, "tIoU": 0.49648056, "tRecall_50": 0.525, "tRecall_75": 0.35625, "tRecall_90": 0.2125, "token_acc": 0.93736443, "epoch": 0.59797913, "global_step/max_steps": "1805/3019", "percentage": "59.79%", "elapsed_time": "6h 46m 49s", "remaining_time": "4h 33m 36s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073948} +{"loss": 0.16412354, "grad_norm": 0.61940235, "learning_rate": 3.65e-06, "tIoU": 0.51166811, "tRecall_50": 0.58125, "tRecall_75": 0.35625, "tRecall_90": 0.175, "token_acc": 0.93437244, "epoch": 0.59963558, "global_step/max_steps": "1810/3019", "percentage": "59.95%", "elapsed_time": "6h 48m 0s", "remaining_time": "4h 32m 31s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073938} +{"loss": 0.1682373, "grad_norm": 0.58546472, "learning_rate": 3.62e-06, "tIoU": 0.51624622, "tRecall_50": 0.58125, "tRecall_75": 0.35, "tRecall_90": 0.21875, "token_acc": 0.93845316, "epoch": 0.60129203, "global_step/max_steps": "1815/3019", "percentage": "60.12%", "elapsed_time": "6h 49m 8s", "remaining_time": "4h 31m 24s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073935} +{"loss": 0.16972656, "grad_norm": 0.74054956, "learning_rate": 3.6e-06, "tIoU": 0.53639045, "tRecall_50": 0.58125, "tRecall_75": 0.4125, "tRecall_90": 0.21875, "token_acc": 0.93637847, "epoch": 0.60294848, "global_step/max_steps": "1820/3019", "percentage": "60.28%", "elapsed_time": "6h 50m 14s", "remaining_time": "4h 30m 16s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073939} +{"loss": 0.16621094, "grad_norm": 0.45874158, "learning_rate": 3.57e-06, "tIoU": 0.55258727, "tRecall_50": 0.61875, "tRecall_75": 0.38125, "tRecall_90": 0.2125, "token_acc": 0.93991881, "epoch": 0.60460494, "global_step/max_steps": "1825/3019", "percentage": "60.45%", "elapsed_time": "6h 51m 23s", "remaining_time": "4h 29m 9s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073936} +{"loss": 0.16411133, "grad_norm": 0.68391132, "learning_rate": 3.55e-06, "tIoU": 0.49982576, "tRecall_50": 0.51875, "tRecall_75": 0.325, "tRecall_90": 0.20625, "token_acc": 0.93404197, "epoch": 0.60626139, "global_step/max_steps": "1830/3019", "percentage": "60.62%", "elapsed_time": "6h 52m 31s", "remaining_time": "4h 28m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073936} +{"loss": 0.1625, "grad_norm": 0.66279316, "learning_rate": 3.52e-06, "tIoU": 0.52807732, "tRecall_50": 0.575, "tRecall_75": 0.375, "tRecall_90": 0.2375, "token_acc": 0.93656412, "epoch": 0.60791784, "global_step/max_steps": "1835/3019", "percentage": "60.78%", "elapsed_time": "6h 53m 37s", "remaining_time": "4h 26m 53s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073939} +{"loss": 0.16345215, "grad_norm": 0.81856173, "learning_rate": 3.49e-06, "tIoU": 0.49967675, "tRecall_50": 0.5375, "tRecall_75": 0.34375, "tRecall_90": 0.2, "token_acc": 0.93907104, "epoch": 0.60957429, "global_step/max_steps": "1840/3019", "percentage": "60.95%", "elapsed_time": "6h 54m 47s", "remaining_time": "4h 25m 47s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073932} +{"loss": 0.15917969, "grad_norm": 0.49641514, "learning_rate": 3.47e-06, "tIoU": 0.54198313, "tRecall_50": 0.58125, "tRecall_75": 0.40625, "tRecall_90": 0.28125, "token_acc": 0.94146474, "epoch": 0.61123074, "global_step/max_steps": "1845/3019", "percentage": "61.11%", "elapsed_time": "6h 55m 53s", "remaining_time": "4h 24m 38s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073938} +{"loss": 0.18828125, "grad_norm": 0.77559626, "learning_rate": 3.44e-06, "tIoU": 0.48210011, "tRecall_50": 0.5125, "tRecall_75": 0.36875, "tRecall_90": 0.175, "token_acc": 0.93287858, "epoch": 0.6128872, "global_step/max_steps": "1850/3019", "percentage": "61.28%", "elapsed_time": "6h 57m 0s", "remaining_time": "4h 23m 30s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07394} +{"loss": 0.1717041, "grad_norm": 0.42528418, "learning_rate": 3.42e-06, "tIoU": 0.50079975, "tRecall_50": 0.5125, "tRecall_75": 0.33125, "tRecall_90": 0.19375, "token_acc": 0.93376588, "epoch": 0.61454365, "global_step/max_steps": "1855/3019", "percentage": "61.44%", "elapsed_time": "6h 58m 8s", "remaining_time": "4h 22m 23s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073937} +{"loss": 0.15889893, "grad_norm": 0.54965156, "learning_rate": 3.39e-06, "tIoU": 0.51122746, "tRecall_50": 0.54375, "tRecall_75": 0.375, "tRecall_90": 0.24375, "token_acc": 0.94066413, "epoch": 0.6162001, "global_step/max_steps": "1860/3019", "percentage": "61.61%", "elapsed_time": "6h 59m 17s", "remaining_time": "4h 21m 16s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073934} +{"loss": 0.1677063, "grad_norm": 0.5841198, "learning_rate": 3.37e-06, "tIoU": 0.51001825, "tRecall_50": 0.55, "tRecall_75": 0.3375, "tRecall_90": 0.2375, "token_acc": 0.93649496, "epoch": 0.61785655, "global_step/max_steps": "1865/3019", "percentage": "61.78%", "elapsed_time": "7h 0m 30s", "remaining_time": "4h 20m 12s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073917} +{"loss": 0.16390381, "grad_norm": 0.85970521, "learning_rate": 3.34e-06, "tIoU": 0.50235624, "tRecall_50": 0.5625, "tRecall_75": 0.3625, "tRecall_90": 0.2, "token_acc": 0.93697135, "epoch": 0.619513, "global_step/max_steps": "1870/3019", "percentage": "61.94%", "elapsed_time": "7h 1m 44s", "remaining_time": "4h 19m 7s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073901} +{"loss": 0.17318115, "grad_norm": 0.4235498, "learning_rate": 3.32e-06, "tIoU": 0.47911148, "tRecall_50": 0.5375, "tRecall_75": 0.3125, "tRecall_90": 0.15625, "token_acc": 0.93409276, "epoch": 0.62116946, "global_step/max_steps": "1875/3019", "percentage": "62.11%", "elapsed_time": "7h 2m 50s", "remaining_time": "4h 17m 59s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073904} +{"loss": 0.17980957, "grad_norm": 0.5873208, "learning_rate": 3.29e-06, "tIoU": 0.50178723, "tRecall_50": 0.55, "tRecall_75": 0.375, "tRecall_90": 0.2375, "token_acc": 0.93414966, "epoch": 0.62282591, "global_step/max_steps": "1880/3019", "percentage": "62.27%", "elapsed_time": "7h 3m 58s", "remaining_time": "4h 16m 51s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073905} +{"loss": 0.16212158, "grad_norm": 0.27464816, "learning_rate": 3.27e-06, "tIoU": 0.53211498, "tRecall_50": 0.58125, "tRecall_75": 0.3875, "tRecall_90": 0.25625, "token_acc": 0.93631622, "epoch": 0.62448236, "global_step/max_steps": "1885/3019", "percentage": "62.44%", "elapsed_time": "7h 5m 3s", "remaining_time": "4h 15m 42s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07391} +{"loss": 0.16231079, "grad_norm": 0.57387424, "learning_rate": 3.24e-06, "tIoU": 0.52013488, "tRecall_50": 0.56875, "tRecall_75": 0.3875, "tRecall_90": 0.23125, "token_acc": 0.94013605, "epoch": 0.62613881, "global_step/max_steps": "1890/3019", "percentage": "62.60%", "elapsed_time": "7h 6m 13s", "remaining_time": "4h 14m 36s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073905} +{"loss": 0.18139648, "grad_norm": 0.44134361, "learning_rate": 3.22e-06, "tIoU": 0.44995375, "tRecall_50": 0.4625, "tRecall_75": 0.325, "tRecall_90": 0.2, "token_acc": 0.93134328, "epoch": 0.62779526, "global_step/max_steps": "1895/3019", "percentage": "62.77%", "elapsed_time": "7h 7m 18s", "remaining_time": "4h 13m 27s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073913} +{"loss": 0.16608887, "grad_norm": 0.49762654, "learning_rate": 3.19e-06, "tIoU": 0.52428379, "tRecall_50": 0.54375, "tRecall_75": 0.375, "tRecall_90": 0.25625, "token_acc": 0.93748286, "epoch": 0.62945171, "global_step/max_steps": "1900/3019", "percentage": "62.93%", "elapsed_time": "7h 8m 26s", "remaining_time": "4h 12m 19s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073911} +{"loss": 0.15696411, "grad_norm": 0.62759578, "learning_rate": 3.17e-06, "tIoU": 0.56292167, "tRecall_50": 0.59375, "tRecall_75": 0.46875, "tRecall_90": 0.3, "token_acc": 0.94511196, "epoch": 0.63110817, "global_step/max_steps": "1905/3019", "percentage": "63.10%", "elapsed_time": "7h 9m 32s", "remaining_time": "4h 11m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073916} +{"loss": 0.17091064, "grad_norm": 0.59663117, "learning_rate": 3.14e-06, "tIoU": 0.55712352, "tRecall_50": 0.6, "tRecall_75": 0.45, "tRecall_90": 0.2625, "token_acc": 0.93542234, "epoch": 0.63276462, "global_step/max_steps": "1910/3019", "percentage": "63.27%", "elapsed_time": "7h 10m 37s", "remaining_time": "4h 10m 2s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073923} +{"loss": 0.16427002, "grad_norm": 0.46508521, "learning_rate": 3.12e-06, "tIoU": 0.50829773, "tRecall_50": 0.55625, "tRecall_75": 0.34375, "tRecall_90": 0.2125, "token_acc": 0.93584494, "epoch": 0.63442107, "global_step/max_steps": "1915/3019", "percentage": "63.43%", "elapsed_time": "7h 11m 46s", "remaining_time": "4h 8m 55s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07392} +{"loss": 0.16798096, "grad_norm": 0.53463346, "learning_rate": 3.09e-06, "tIoU": 0.49363961, "tRecall_50": 0.525, "tRecall_75": 0.375, "tRecall_90": 0.26875, "token_acc": 0.93881453, "epoch": 0.63607752, "global_step/max_steps": "1920/3019", "percentage": "63.60%", "elapsed_time": "7h 12m 54s", "remaining_time": "4h 7m 47s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073919} +{"loss": 0.17302246, "grad_norm": 0.39497977, "learning_rate": 3.07e-06, "tIoU": 0.53191533, "tRecall_50": 0.58125, "tRecall_75": 0.40625, "tRecall_90": 0.2625, "token_acc": 0.93563281, "epoch": 0.63773397, "global_step/max_steps": "1925/3019", "percentage": "63.76%", "elapsed_time": "7h 13m 58s", "remaining_time": "4h 6m 38s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073929} +{"loss": 0.18037109, "grad_norm": 0.7183004, "learning_rate": 3.04e-06, "tIoU": 0.47457814, "tRecall_50": 0.51875, "tRecall_75": 0.3, "tRecall_90": 0.16875, "token_acc": 0.93109106, "epoch": 0.63939043, "global_step/max_steps": "1930/3019", "percentage": "63.93%", "elapsed_time": "7h 15m 6s", "remaining_time": "4h 5m 30s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073929} +{"loss": 0.16229248, "grad_norm": 0.44636396, "learning_rate": 3.02e-06, "tIoU": 0.52823486, "tRecall_50": 0.55625, "tRecall_75": 0.36875, "tRecall_90": 0.24375, "token_acc": 0.93813028, "epoch": 0.64104688, "global_step/max_steps": "1935/3019", "percentage": "64.09%", "elapsed_time": "7h 16m 8s", "remaining_time": "4h 4m 20s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073943} +{"loss": 0.17168579, "grad_norm": 0.81296653, "learning_rate": 2.99e-06, "tIoU": 0.50463524, "tRecall_50": 0.5375, "tRecall_75": 0.375, "tRecall_90": 0.25625, "token_acc": 0.93807713, "epoch": 0.64270333, "global_step/max_steps": "1940/3019", "percentage": "64.26%", "elapsed_time": "7h 17m 14s", "remaining_time": "4h 3m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073947} +{"loss": 0.16398315, "grad_norm": 0.58241111, "learning_rate": 2.97e-06, "tIoU": 0.55755887, "tRecall_50": 0.60625, "tRecall_75": 0.40625, "tRecall_90": 0.25625, "token_acc": 0.94114426, "epoch": 0.64435978, "global_step/max_steps": "1945/3019", "percentage": "64.43%", "elapsed_time": "7h 18m 23s", "remaining_time": "4h 2m 4s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073946} +{"loss": 0.15794678, "grad_norm": 0.75489819, "learning_rate": 2.94e-06, "tIoU": 0.51483932, "tRecall_50": 0.55, "tRecall_75": 0.425, "tRecall_90": 0.23125, "token_acc": 0.94045677, "epoch": 0.64601623, "global_step/max_steps": "1950/3019", "percentage": "64.59%", "elapsed_time": "7h 19m 29s", "remaining_time": "4h 0m 55s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073949} +{"loss": 0.16236572, "grad_norm": 0.54922247, "learning_rate": 2.92e-06, "tIoU": 0.52466359, "tRecall_50": 0.575, "tRecall_75": 0.4, "tRecall_90": 0.20625, "token_acc": 0.940586, "epoch": 0.64767269, "global_step/max_steps": "1955/3019", "percentage": "64.76%", "elapsed_time": "7h 20m 33s", "remaining_time": "3h 59m 46s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073958} +{"loss": 0.15792236, "grad_norm": 0.59774548, "learning_rate": 2.9e-06, "tIoU": 0.51488633, "tRecall_50": 0.56875, "tRecall_75": 0.3875, "tRecall_90": 0.23125, "token_acc": 0.94052146, "epoch": 0.64932914, "global_step/max_steps": "1960/3019", "percentage": "64.92%", "elapsed_time": "7h 21m 41s", "remaining_time": "3h 58m 38s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073959} +{"loss": 0.15865479, "grad_norm": 0.40176132, "learning_rate": 2.87e-06, "tIoU": 0.5105423, "tRecall_50": 0.5375, "tRecall_75": 0.35625, "tRecall_90": 0.2625, "token_acc": 0.93935273, "epoch": 0.65098559, "global_step/max_steps": "1965/3019", "percentage": "65.09%", "elapsed_time": "7h 22m 46s", "remaining_time": "3h 57m 30s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073965} +{"loss": 0.15938721, "grad_norm": 0.64463896, "learning_rate": 2.85e-06, "tIoU": 0.52905402, "tRecall_50": 0.5625, "tRecall_75": 0.36875, "tRecall_90": 0.275, "token_acc": 0.94288822, "epoch": 0.65264204, "global_step/max_steps": "1970/3019", "percentage": "65.25%", "elapsed_time": "7h 23m 52s", "remaining_time": "3h 56m 21s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073969} +{"loss": 0.16928711, "grad_norm": 0.41462159, "learning_rate": 2.82e-06, "tIoU": 0.50186563, "tRecall_50": 0.4875, "tRecall_75": 0.35625, "tRecall_90": 0.25625, "token_acc": 0.94040816, "epoch": 0.65429849, "global_step/max_steps": "1975/3019", "percentage": "65.42%", "elapsed_time": "7h 25m 2s", "remaining_time": "3h 55m 15s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073963} +{"loss": 0.16522217, "grad_norm": 0.95643109, "learning_rate": 2.8e-06, "tIoU": 0.5413733, "tRecall_50": 0.6, "tRecall_75": 0.39375, "tRecall_90": 0.2625, "token_acc": 0.93843639, "epoch": 0.65595494, "global_step/max_steps": "1980/3019", "percentage": "65.58%", "elapsed_time": "7h 26m 12s", "remaining_time": "3h 54m 8s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073957} +{"loss": 0.16549072, "grad_norm": 0.60067457, "learning_rate": 2.77e-06, "tIoU": 0.54275556, "tRecall_50": 0.6125, "tRecall_75": 0.3875, "tRecall_90": 0.275, "token_acc": 0.93953362, "epoch": 0.6576114, "global_step/max_steps": "1985/3019", "percentage": "65.75%", "elapsed_time": "7h 27m 22s", "remaining_time": "3h 53m 2s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073949} +{"loss": 0.15649414, "grad_norm": 0.6821838, "learning_rate": 2.75e-06, "tIoU": 0.54003606, "tRecall_50": 0.5625, "tRecall_75": 0.4125, "tRecall_90": 0.28125, "token_acc": 0.944869, "epoch": 0.65926785, "global_step/max_steps": "1990/3019", "percentage": "65.92%", "elapsed_time": "7h 28m 33s", "remaining_time": "3h 51m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073942} +{"loss": 0.15616455, "grad_norm": 0.51665086, "learning_rate": 2.73e-06, "tIoU": 0.53555565, "tRecall_50": 0.60625, "tRecall_75": 0.36875, "tRecall_90": 0.21875, "token_acc": 0.94061303, "epoch": 0.6609243, "global_step/max_steps": "1995/3019", "percentage": "66.08%", "elapsed_time": "7h 29m 38s", "remaining_time": "3h 50m 47s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073948} +{"loss": 0.17166748, "grad_norm": 0.70358044, "learning_rate": 2.7e-06, "tIoU": 0.50961804, "tRecall_50": 0.54375, "tRecall_75": 0.3625, "tRecall_90": 0.20625, "token_acc": 0.93581724, "epoch": 0.66258075, "global_step/max_steps": "2000/3019", "percentage": "66.25%", "elapsed_time": "7h 30m 51s", "remaining_time": "3h 49m 42s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073934} +{"loss": 0.16157837, "grad_norm": 0.5203203, "learning_rate": 2.68e-06, "tIoU": 0.52877412, "tRecall_50": 0.5625, "tRecall_75": 0.35625, "tRecall_90": 0.23125, "token_acc": 0.94009343, "epoch": 0.6642372, "global_step/max_steps": "2005/3019", "percentage": "66.41%", "elapsed_time": "7h 32m 1s", "remaining_time": "3h 48m 36s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073927} +{"loss": 0.168927, "grad_norm": 0.83931428, "learning_rate": 2.65e-06, "tIoU": 0.51892419, "tRecall_50": 0.55625, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.94003286, "epoch": 0.66589366, "global_step/max_steps": "2010/3019", "percentage": "66.58%", "elapsed_time": "7h 33m 13s", "remaining_time": "3h 47m 30s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073914} +{"loss": 0.16002197, "grad_norm": 0.47614089, "learning_rate": 2.63e-06, "tIoU": 0.49315405, "tRecall_50": 0.55, "tRecall_75": 0.35625, "tRecall_90": 0.2625, "token_acc": 0.93691972, "epoch": 0.66755011, "global_step/max_steps": "2015/3019", "percentage": "66.74%", "elapsed_time": "7h 34m 24s", "remaining_time": "3h 46m 24s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073906} +{"loss": 0.16381836, "grad_norm": 0.63881332, "learning_rate": 2.61e-06, "tIoU": 0.54713607, "tRecall_50": 0.5875, "tRecall_75": 0.4, "tRecall_90": 0.26875, "token_acc": 0.94250068, "epoch": 0.66920656, "global_step/max_steps": "2020/3019", "percentage": "66.91%", "elapsed_time": "7h 35m 30s", "remaining_time": "3h 45m 16s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073909} +{"loss": 0.17288818, "grad_norm": 0.47899044, "learning_rate": 2.58e-06, "tIoU": 0.4950345, "tRecall_50": 0.53125, "tRecall_75": 0.33125, "tRecall_90": 0.2625, "token_acc": 0.93830194, "epoch": 0.67086301, "global_step/max_steps": "2025/3019", "percentage": "67.08%", "elapsed_time": "7h 36m 40s", "remaining_time": "3h 44m 9s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073904} +{"loss": 0.16365967, "grad_norm": 1.09099293, "learning_rate": 2.56e-06, "tIoU": 0.52656164, "tRecall_50": 0.56875, "tRecall_75": 0.4, "tRecall_90": 0.25, "token_acc": 0.94149809, "epoch": 0.67251946, "global_step/max_steps": "2030/3019", "percentage": "67.24%", "elapsed_time": "7h 37m 52s", "remaining_time": "3h 43m 4s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073892} +{"loss": 0.15804443, "grad_norm": 0.57685786, "learning_rate": 2.54e-06, "tIoU": 0.50101748, "tRecall_50": 0.53125, "tRecall_75": 0.4, "tRecall_90": 0.23125, "token_acc": 0.93647764, "epoch": 0.67417592, "global_step/max_steps": "2035/3019", "percentage": "67.41%", "elapsed_time": "7h 38m 56s", "remaining_time": "3h 41m 55s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073901} +{"loss": 0.16638184, "grad_norm": 0.8344062, "learning_rate": 2.51e-06, "tIoU": 0.5199985, "tRecall_50": 0.50625, "tRecall_75": 0.40625, "tRecall_90": 0.28125, "token_acc": 0.9410802, "epoch": 0.67583237, "global_step/max_steps": "2040/3019", "percentage": "67.57%", "elapsed_time": "7h 40m 5s", "remaining_time": "3h 40m 48s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073898} +{"loss": 0.16413574, "grad_norm": 0.74369222, "learning_rate": 2.49e-06, "tIoU": 0.47923422, "tRecall_50": 0.5125, "tRecall_75": 0.3375, "tRecall_90": 0.225, "token_acc": 0.93758517, "epoch": 0.67748882, "global_step/max_steps": "2045/3019", "percentage": "67.74%", "elapsed_time": "7h 41m 16s", "remaining_time": "3h 39m 41s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073889} +{"loss": 0.1564209, "grad_norm": 0.56531888, "learning_rate": 2.47e-06, "tIoU": 0.55842662, "tRecall_50": 0.6, "tRecall_75": 0.4625, "tRecall_90": 0.2875, "token_acc": 0.94160982, "epoch": 0.67914527, "global_step/max_steps": "2050/3019", "percentage": "67.90%", "elapsed_time": "7h 42m 21s", "remaining_time": "3h 38m 33s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073895} +{"loss": 0.17310791, "grad_norm": 0.57359064, "learning_rate": 2.44e-06, "tIoU": 0.4905633, "tRecall_50": 0.54375, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.93526544, "epoch": 0.68080172, "global_step/max_steps": "2055/3019", "percentage": "68.07%", "elapsed_time": "7h 43m 25s", "remaining_time": "3h 37m 23s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073906} +{"loss": 0.16672363, "grad_norm": 0.73452705, "learning_rate": 2.42e-06, "tIoU": 0.51104608, "tRecall_50": 0.56875, "tRecall_75": 0.35, "tRecall_90": 0.2, "token_acc": 0.93536951, "epoch": 0.68245817, "global_step/max_steps": "2060/3019", "percentage": "68.23%", "elapsed_time": "7h 44m 35s", "remaining_time": "3h 36m 17s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.0739} +{"loss": 0.15985107, "grad_norm": 1.27823746, "learning_rate": 2.4e-06, "tIoU": 0.56404977, "tRecall_50": 0.60625, "tRecall_75": 0.44375, "tRecall_90": 0.2625, "token_acc": 0.94175554, "epoch": 0.68411463, "global_step/max_steps": "2065/3019", "percentage": "68.40%", "elapsed_time": "7h 45m 44s", "remaining_time": "3h 35m 9s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073898} +{"loss": 0.16499023, "grad_norm": 0.58060801, "learning_rate": 2.38e-06, "tIoU": 0.49380932, "tRecall_50": 0.54375, "tRecall_75": 0.3625, "tRecall_90": 0.24375, "token_acc": 0.9363092, "epoch": 0.68577108, "global_step/max_steps": "2070/3019", "percentage": "68.57%", "elapsed_time": "7h 46m 51s", "remaining_time": "3h 34m 2s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073898} +{"loss": 0.16665039, "grad_norm": 0.59931004, "learning_rate": 2.35e-06, "tIoU": 0.52099314, "tRecall_50": 0.575, "tRecall_75": 0.38125, "tRecall_90": 0.2375, "token_acc": 0.93980477, "epoch": 0.68742753, "global_step/max_steps": "2075/3019", "percentage": "68.73%", "elapsed_time": "7h 48m 2s", "remaining_time": "3h 32m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073888} +{"loss": 0.16470947, "grad_norm": 0.63605988, "learning_rate": 2.33e-06, "tIoU": 0.50229959, "tRecall_50": 0.50625, "tRecall_75": 0.36875, "tRecall_90": 0.26875, "token_acc": 0.93657736, "epoch": 0.68908398, "global_step/max_steps": "2080/3019", "percentage": "68.90%", "elapsed_time": "7h 49m 9s", "remaining_time": "3h 31m 47s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073892} +{"loss": 0.16615601, "grad_norm": 0.62258261, "learning_rate": 2.31e-06, "tIoU": 0.53280134, "tRecall_50": 0.55625, "tRecall_75": 0.4, "tRecall_90": 0.26875, "token_acc": 0.94109626, "epoch": 0.69074043, "global_step/max_steps": "2085/3019", "percentage": "69.06%", "elapsed_time": "7h 50m 18s", "remaining_time": "3h 30m 40s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073888} +{"loss": 0.16364746, "grad_norm": 0.54338008, "learning_rate": 2.28e-06, "tIoU": 0.49683642, "tRecall_50": 0.5375, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.938549, "epoch": 0.69239689, "global_step/max_steps": "2090/3019", "percentage": "69.23%", "elapsed_time": "7h 51m 28s", "remaining_time": "3h 29m 34s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07388} +{"loss": 0.16856079, "grad_norm": 0.67706835, "learning_rate": 2.26e-06, "tIoU": 0.4816958, "tRecall_50": 0.5125, "tRecall_75": 0.35625, "tRecall_90": 0.2375, "token_acc": 0.93542234, "epoch": 0.69405334, "global_step/max_steps": "2095/3019", "percentage": "69.39%", "elapsed_time": "7h 52m 34s", "remaining_time": "3h 28m 25s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073886} +{"loss": 0.16677246, "grad_norm": 0.49540141, "learning_rate": 2.24e-06, "tIoU": 0.52496767, "tRecall_50": 0.55625, "tRecall_75": 0.4, "tRecall_90": 0.23125, "token_acc": 0.94042437, "epoch": 0.69570979, "global_step/max_steps": "2100/3019", "percentage": "69.56%", "elapsed_time": "7h 53m 41s", "remaining_time": "3h 27m 17s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073889} +{"loss": 0.16582031, "grad_norm": 0.50589752, "learning_rate": 2.22e-06, "tIoU": 0.52024808, "tRecall_50": 0.54375, "tRecall_75": 0.38125, "tRecall_90": 0.25, "token_acc": 0.94000545, "epoch": 0.69736624, "global_step/max_steps": "2105/3019", "percentage": "69.73%", "elapsed_time": "7h 54m 45s", "remaining_time": "3h 26m 8s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073897} +{"loss": 0.1644043, "grad_norm": 0.37972459, "learning_rate": 2.2e-06, "tIoU": 0.51444387, "tRecall_50": 0.54375, "tRecall_75": 0.4125, "tRecall_90": 0.2625, "token_acc": 0.9379085, "epoch": 0.69902269, "global_step/max_steps": "2110/3019", "percentage": "69.89%", "elapsed_time": "7h 55m 53s", "remaining_time": "3h 25m 0s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073898} +{"loss": 0.16228027, "grad_norm": 0.60593903, "learning_rate": 2.17e-06, "tIoU": 0.50943608, "tRecall_50": 0.5625, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.93659864, "epoch": 0.70067915, "global_step/max_steps": "2115/3019", "percentage": "70.06%", "elapsed_time": "7h 56m 59s", "remaining_time": "3h 23m 52s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.0739} +{"loss": 0.16075439, "grad_norm": 0.7819739, "learning_rate": 2.15e-06, "tIoU": 0.58101783, "tRecall_50": 0.6125, "tRecall_75": 0.43125, "tRecall_90": 0.3125, "token_acc": 0.94247546, "epoch": 0.7023356, "global_step/max_steps": "2120/3019", "percentage": "70.22%", "elapsed_time": "7h 58m 11s", "remaining_time": "3h 22m 46s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073891} +{"loss": 0.16397705, "grad_norm": 0.54920578, "learning_rate": 2.13e-06, "tIoU": 0.53585353, "tRecall_50": 0.575, "tRecall_75": 0.39375, "tRecall_90": 0.225, "token_acc": 0.93941048, "epoch": 0.70399205, "global_step/max_steps": "2125/3019", "percentage": "70.39%", "elapsed_time": "7h 59m 16s", "remaining_time": "3h 21m 38s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073896} +{"loss": 0.16303711, "grad_norm": 0.50172567, "learning_rate": 2.11e-06, "tIoU": 0.52140121, "tRecall_50": 0.60625, "tRecall_75": 0.31875, "tRecall_90": 0.2, "token_acc": 0.93649496, "epoch": 0.7056485, "global_step/max_steps": "2130/3019", "percentage": "70.55%", "elapsed_time": "8h 0m 25s", "remaining_time": "3h 20m 31s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073892} +{"loss": 0.17164307, "grad_norm": 0.60925078, "learning_rate": 2.09e-06, "tIoU": 0.50379291, "tRecall_50": 0.58125, "tRecall_75": 0.33125, "tRecall_90": 0.2, "token_acc": 0.93675027, "epoch": 0.70730495, "global_step/max_steps": "2135/3019", "percentage": "70.72%", "elapsed_time": "8h 1m 31s", "remaining_time": "3h 19m 22s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073896} +{"loss": 0.16177979, "grad_norm": 0.53455752, "learning_rate": 2.06e-06, "tIoU": 0.50682273, "tRecall_50": 0.55, "tRecall_75": 0.39375, "tRecall_90": 0.20625, "token_acc": 0.94042437, "epoch": 0.7089614, "global_step/max_steps": "2140/3019", "percentage": "70.88%", "elapsed_time": "8h 2m 42s", "remaining_time": "3h 18m 16s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07389} +{"loss": 0.16123657, "grad_norm": 0.83791137, "learning_rate": 2.04e-06, "tIoU": 0.51363969, "tRecall_50": 0.51875, "tRecall_75": 0.35625, "tRecall_90": 0.2375, "token_acc": 0.9387088, "epoch": 0.71061786, "global_step/max_steps": "2145/3019", "percentage": "71.05%", "elapsed_time": "8h 3m 49s", "remaining_time": "3h 17m 8s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073889} +{"loss": 0.15539551, "grad_norm": 0.50986147, "learning_rate": 2.02e-06, "tIoU": 0.57358451, "tRecall_50": 0.6375, "tRecall_75": 0.4125, "tRecall_90": 0.3, "token_acc": 0.94132029, "epoch": 0.71227431, "global_step/max_steps": "2150/3019", "percentage": "71.22%", "elapsed_time": "8h 4m 59s", "remaining_time": "3h 16m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073885} +{"loss": 0.16419678, "grad_norm": 0.75286472, "learning_rate": 2e-06, "tIoU": 0.53306808, "tRecall_50": 0.5625, "tRecall_75": 0.4, "tRecall_90": 0.21875, "token_acc": 0.93872549, "epoch": 0.71393076, "global_step/max_steps": "2155/3019", "percentage": "71.38%", "elapsed_time": "8h 6m 15s", "remaining_time": "3h 14m 57s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073863} +{"loss": 0.16022949, "grad_norm": 0.47473061, "learning_rate": 1.98e-06, "tIoU": 0.52057805, "tRecall_50": 0.58125, "tRecall_75": 0.3875, "tRecall_90": 0.2, "token_acc": 0.9394844, "epoch": 0.71558721, "global_step/max_steps": "2160/3019", "percentage": "71.55%", "elapsed_time": "8h 7m 22s", "remaining_time": "3h 13m 49s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073866} +{"loss": 0.17667236, "grad_norm": 0.95846772, "learning_rate": 1.96e-06, "tIoU": 0.47919407, "tRecall_50": 0.50625, "tRecall_75": 0.38125, "tRecall_90": 0.21875, "token_acc": 0.93527508, "epoch": 0.71724366, "global_step/max_steps": "2165/3019", "percentage": "71.71%", "elapsed_time": "8h 8m 33s", "remaining_time": "3h 12m 43s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073856} +{"loss": 0.14588013, "grad_norm": 0.56558079, "learning_rate": 1.93e-06, "tIoU": 0.58003821, "tRecall_50": 0.6125, "tRecall_75": 0.45, "tRecall_90": 0.28125, "token_acc": 0.94491642, "epoch": 0.71890012, "global_step/max_steps": "2170/3019", "percentage": "71.88%", "elapsed_time": "8h 9m 44s", "remaining_time": "3h 11m 36s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073848} +{"loss": 0.15904541, "grad_norm": 0.48551893, "learning_rate": 1.91e-06, "tIoU": 0.57859669, "tRecall_50": 0.6375, "tRecall_75": 0.4625, "tRecall_90": 0.31875, "token_acc": 0.94436832, "epoch": 0.72055657, "global_step/max_steps": "2175/3019", "percentage": "72.04%", "elapsed_time": "8h 10m 52s", "remaining_time": "3h 10m 29s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073847} +{"loss": 0.16480713, "grad_norm": 0.42149362, "learning_rate": 1.89e-06, "tIoU": 0.54441257, "tRecall_50": 0.56875, "tRecall_75": 0.40625, "tRecall_90": 0.2625, "token_acc": 0.93756786, "epoch": 0.72221302, "global_step/max_steps": "2180/3019", "percentage": "72.21%", "elapsed_time": "8h 11m 56s", "remaining_time": "3h 9m 19s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073856} +{"loss": 0.15125122, "grad_norm": 0.8240121, "learning_rate": 1.87e-06, "tIoU": 0.55659504, "tRecall_50": 0.6, "tRecall_75": 0.41875, "tRecall_90": 0.2875, "token_acc": 0.94500135, "epoch": 0.72386947, "global_step/max_steps": "2185/3019", "percentage": "72.37%", "elapsed_time": "8h 13m 1s", "remaining_time": "3h 8m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073863} +{"loss": 0.16071777, "grad_norm": 0.41935021, "learning_rate": 1.85e-06, "tIoU": 0.54103723, "tRecall_50": 0.575, "tRecall_75": 0.40625, "tRecall_90": 0.30625, "token_acc": 0.94298127, "epoch": 0.72552592, "global_step/max_steps": "2190/3019", "percentage": "72.54%", "elapsed_time": "8h 14m 4s", "remaining_time": "3h 7m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073875} +{"loss": 0.16315918, "grad_norm": 0.61023593, "learning_rate": 1.83e-06, "tIoU": 0.49234715, "tRecall_50": 0.525, "tRecall_75": 0.3375, "tRecall_90": 0.25, "token_acc": 0.93695175, "epoch": 0.72718238, "global_step/max_steps": "2195/3019", "percentage": "72.71%", "elapsed_time": "8h 15m 8s", "remaining_time": "3h 5m 52s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073884} +{"loss": 0.13913574, "grad_norm": 0.59919435, "learning_rate": 1.81e-06, "tIoU": 0.59030833, "tRecall_50": 0.61875, "tRecall_75": 0.49375, "tRecall_90": 0.30625, "token_acc": 0.94623362, "epoch": 0.72883883, "global_step/max_steps": "2200/3019", "percentage": "72.87%", "elapsed_time": "8h 16m 14s", "remaining_time": "3h 4m 44s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07389} +{"loss": 0.15266113, "grad_norm": 0.66858393, "learning_rate": 1.79e-06, "tIoU": 0.58086934, "tRecall_50": 0.60625, "tRecall_75": 0.4375, "tRecall_90": 0.3375, "token_acc": 0.94406361, "epoch": 0.73049528, "global_step/max_steps": "2205/3019", "percentage": "73.04%", "elapsed_time": "8h 17m 20s", "remaining_time": "3h 3m 36s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073892} +{"loss": 0.16241455, "grad_norm": 0.52423912, "learning_rate": 1.77e-06, "tIoU": 0.52096596, "tRecall_50": 0.5625, "tRecall_75": 0.39375, "tRecall_90": 0.24375, "token_acc": 0.93731262, "epoch": 0.73215173, "global_step/max_steps": "2210/3019", "percentage": "73.20%", "elapsed_time": "8h 18m 31s", "remaining_time": "3h 2m 29s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073885} +{"loss": 0.16380615, "grad_norm": 0.62338924, "learning_rate": 1.75e-06, "tIoU": 0.51739268, "tRecall_50": 0.55625, "tRecall_75": 0.3875, "tRecall_90": 0.225, "token_acc": 0.94004927, "epoch": 0.73380818, "global_step/max_steps": "2215/3019", "percentage": "73.37%", "elapsed_time": "8h 19m 41s", "remaining_time": "3h 1m 22s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073879} +{"loss": 0.16237793, "grad_norm": 0.89424825, "learning_rate": 1.73e-06, "tIoU": 0.55924439, "tRecall_50": 0.6, "tRecall_75": 0.4, "tRecall_90": 0.28125, "token_acc": 0.94350437, "epoch": 0.73546463, "global_step/max_steps": "2220/3019", "percentage": "73.53%", "elapsed_time": "8h 20m 51s", "remaining_time": "3h 0m 15s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073874} +{"loss": 0.17415161, "grad_norm": 0.73215139, "learning_rate": 1.71e-06, "tIoU": 0.45665102, "tRecall_50": 0.4875, "tRecall_75": 0.33125, "tRecall_90": 0.14375, "token_acc": 0.93288043, "epoch": 0.73712109, "global_step/max_steps": "2225/3019", "percentage": "73.70%", "elapsed_time": "8h 21m 53s", "remaining_time": "2h 59m 6s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073887} +{"loss": 0.14980469, "grad_norm": 0.65495372, "learning_rate": 1.69e-06, "tIoU": 0.55599146, "tRecall_50": 0.59375, "tRecall_75": 0.41875, "tRecall_90": 0.25, "token_acc": 0.94194962, "epoch": 0.73877754, "global_step/max_steps": "2230/3019", "percentage": "73.87%", "elapsed_time": "8h 22m 56s", "remaining_time": "2h 57m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073899} +{"loss": 0.16259766, "grad_norm": 0.62229824, "learning_rate": 1.67e-06, "tIoU": 0.51217299, "tRecall_50": 0.50625, "tRecall_75": 0.35625, "tRecall_90": 0.2375, "token_acc": 0.93518519, "epoch": 0.74043399, "global_step/max_steps": "2235/3019", "percentage": "74.03%", "elapsed_time": "8h 23m 58s", "remaining_time": "2h 56m 47s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073912} +{"loss": 0.16424561, "grad_norm": 0.65248609, "learning_rate": 1.65e-06, "tIoU": 0.50481267, "tRecall_50": 0.54375, "tRecall_75": 0.35625, "tRecall_90": 0.20625, "token_acc": 0.93712737, "epoch": 0.74209044, "global_step/max_steps": "2240/3019", "percentage": "74.20%", "elapsed_time": "8h 25m 1s", "remaining_time": "2h 55m 38s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073923} +{"loss": 0.16589966, "grad_norm": 0.65702492, "learning_rate": 1.63e-06, "tIoU": 0.52776899, "tRecall_50": 0.55625, "tRecall_75": 0.3625, "tRecall_90": 0.25, "token_acc": 0.93894794, "epoch": 0.74374689, "global_step/max_steps": "2245/3019", "percentage": "74.36%", "elapsed_time": "8h 26m 5s", "remaining_time": "2h 54m 29s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073931} +{"loss": 0.15498047, "grad_norm": 0.57343376, "learning_rate": 1.61e-06, "tIoU": 0.56350703, "tRecall_50": 0.625, "tRecall_75": 0.475, "tRecall_90": 0.2875, "token_acc": 0.94114441, "epoch": 0.74540335, "global_step/max_steps": "2250/3019", "percentage": "74.53%", "elapsed_time": "8h 27m 10s", "remaining_time": "2h 53m 20s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07394} +{"loss": 0.14813843, "grad_norm": 0.54294187, "learning_rate": 1.59e-06, "tIoU": 0.53751173, "tRecall_50": 0.575, "tRecall_75": 0.40625, "tRecall_90": 0.24375, "token_acc": 0.94191437, "epoch": 0.7470598, "global_step/max_steps": "2255/3019", "percentage": "74.69%", "elapsed_time": "8h 28m 13s", "remaining_time": "2h 52m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07395} +{"loss": 0.17375488, "grad_norm": 0.48230994, "learning_rate": 1.57e-06, "tIoU": 0.4863481, "tRecall_50": 0.51875, "tRecall_75": 0.3375, "tRecall_90": 0.19375, "token_acc": 0.93442623, "epoch": 0.74871625, "global_step/max_steps": "2260/3019", "percentage": "74.86%", "elapsed_time": "8h 29m 20s", "remaining_time": "2h 51m 3s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073951} +{"loss": 0.15982056, "grad_norm": 0.57643539, "learning_rate": 1.55e-06, "tIoU": 0.51883368, "tRecall_50": 0.5375, "tRecall_75": 0.33125, "tRecall_90": 0.25, "token_acc": 0.94165758, "epoch": 0.7503727, "global_step/max_steps": "2265/3019", "percentage": "75.02%", "elapsed_time": "8h 30m 29s", "remaining_time": "2h 49m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073949} +{"loss": 0.16252441, "grad_norm": 0.72380298, "learning_rate": 1.53e-06, "tIoU": 0.55865102, "tRecall_50": 0.575, "tRecall_75": 0.36875, "tRecall_90": 0.25625, "token_acc": 0.93950081, "epoch": 0.75202915, "global_step/max_steps": "2270/3019", "percentage": "75.19%", "elapsed_time": "8h 31m 37s", "remaining_time": "2h 48m 48s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073947} +{"loss": 0.16297607, "grad_norm": 0.62636238, "learning_rate": 1.51e-06, "tIoU": 0.53618298, "tRecall_50": 0.575, "tRecall_75": 0.375, "tRecall_90": 0.26875, "token_acc": 0.94031071, "epoch": 0.75368561, "global_step/max_steps": "2275/3019", "percentage": "75.36%", "elapsed_time": "8h 32m 44s", "remaining_time": "2h 47m 41s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073949} +{"loss": 0.17288818, "grad_norm": 0.61710787, "learning_rate": 1.49e-06, "tIoU": 0.51065063, "tRecall_50": 0.525, "tRecall_75": 0.3875, "tRecall_90": 0.25625, "token_acc": 0.933134, "epoch": 0.75534206, "global_step/max_steps": "2280/3019", "percentage": "75.52%", "elapsed_time": "8h 33m 53s", "remaining_time": "2h 46m 33s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073945} +{"loss": 0.15936279, "grad_norm": 0.55264854, "learning_rate": 1.47e-06, "tIoU": 0.51048789, "tRecall_50": 0.525, "tRecall_75": 0.3875, "tRecall_90": 0.23125, "token_acc": 0.93816399, "epoch": 0.75699851, "global_step/max_steps": "2285/3019", "percentage": "75.69%", "elapsed_time": "8h 35m 1s", "remaining_time": "2h 45m 26s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073944} +{"loss": 0.14937744, "grad_norm": 0.71885973, "learning_rate": 1.45e-06, "tIoU": 0.55138855, "tRecall_50": 0.63125, "tRecall_75": 0.4125, "tRecall_90": 0.2625, "token_acc": 0.94575536, "epoch": 0.75865496, "global_step/max_steps": "2290/3019", "percentage": "75.85%", "elapsed_time": "8h 36m 8s", "remaining_time": "2h 44m 18s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073946} +{"loss": 0.16483154, "grad_norm": 0.5185532, "learning_rate": 1.43e-06, "tIoU": 0.53777814, "tRecall_50": 0.5875, "tRecall_75": 0.43125, "tRecall_90": 0.275, "token_acc": 0.94104744, "epoch": 0.76031141, "global_step/max_steps": "2295/3019", "percentage": "76.02%", "elapsed_time": "8h 37m 16s", "remaining_time": "2h 43m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073944} +{"loss": 0.15621948, "grad_norm": 1.12463951, "learning_rate": 1.42e-06, "tIoU": 0.53599548, "tRecall_50": 0.59375, "tRecall_75": 0.3875, "tRecall_90": 0.26875, "token_acc": 0.94221859, "epoch": 0.76196786, "global_step/max_steps": "2300/3019", "percentage": "76.18%", "elapsed_time": "8h 38m 24s", "remaining_time": "2h 42m 3s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073945} +{"loss": 0.16826172, "grad_norm": 0.61187953, "learning_rate": 1.4e-06, "tIoU": 0.52061787, "tRecall_50": 0.5625, "tRecall_75": 0.3875, "tRecall_90": 0.2625, "token_acc": 0.93809654, "epoch": 0.76362432, "global_step/max_steps": "2305/3019", "percentage": "76.35%", "elapsed_time": "8h 39m 28s", "remaining_time": "2h 40m 54s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073953} +{"loss": 0.1671875, "grad_norm": 0.68577313, "learning_rate": 1.38e-06, "tIoU": 0.51814308, "tRecall_50": 0.55625, "tRecall_75": 0.35625, "tRecall_90": 0.2125, "token_acc": 0.93780554, "epoch": 0.76528077, "global_step/max_steps": "2310/3019", "percentage": "76.52%", "elapsed_time": "8h 40m 34s", "remaining_time": "2h 39m 46s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073957} +{"loss": 0.15540771, "grad_norm": 0.5392676, "learning_rate": 1.36e-06, "tIoU": 0.54766817, "tRecall_50": 0.6125, "tRecall_75": 0.3875, "tRecall_90": 0.24375, "token_acc": 0.93852459, "epoch": 0.76693722, "global_step/max_steps": "2315/3019", "percentage": "76.68%", "elapsed_time": "8h 41m 39s", "remaining_time": "2h 38m 38s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073962} +{"loss": 0.16782837, "grad_norm": 0.63624179, "learning_rate": 1.34e-06, "tIoU": 0.524962, "tRecall_50": 0.56875, "tRecall_75": 0.39375, "tRecall_90": 0.225, "token_acc": 0.93997827, "epoch": 0.76859367, "global_step/max_steps": "2320/3019", "percentage": "76.85%", "elapsed_time": "8h 42m 46s", "remaining_time": "2h 37m 30s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073965} +{"loss": 0.15559082, "grad_norm": 0.76028752, "learning_rate": 1.32e-06, "tIoU": 0.56043906, "tRecall_50": 0.6125, "tRecall_75": 0.40625, "tRecall_90": 0.2875, "token_acc": 0.94409599, "epoch": 0.77025012, "global_step/max_steps": "2325/3019", "percentage": "77.01%", "elapsed_time": "8h 43m 55s", "remaining_time": "2h 36m 23s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07396} +{"loss": 0.15029297, "grad_norm": 0.63733929, "learning_rate": 1.31e-06, "tIoU": 0.51829, "tRecall_50": 0.55, "tRecall_75": 0.38125, "tRecall_90": 0.2375, "token_acc": 0.94082356, "epoch": 0.77190658, "global_step/max_steps": "2330/3019", "percentage": "77.18%", "elapsed_time": "8h 45m 4s", "remaining_time": "2h 35m 16s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073957} +{"loss": 0.15554199, "grad_norm": 0.74773151, "learning_rate": 1.29e-06, "tIoU": 0.51711793, "tRecall_50": 0.53125, "tRecall_75": 0.35, "tRecall_90": 0.20625, "token_acc": 0.94026609, "epoch": 0.77356303, "global_step/max_steps": "2335/3019", "percentage": "77.34%", "elapsed_time": "8h 46m 14s", "remaining_time": "2h 34m 9s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073951} +{"loss": 0.15079346, "grad_norm": 0.78949451, "learning_rate": 1.27e-06, "tIoU": 0.5262136, "tRecall_50": 0.56875, "tRecall_75": 0.36875, "tRecall_90": 0.2375, "token_acc": 0.94116037, "epoch": 0.77521948, "global_step/max_steps": "2340/3019", "percentage": "77.51%", "elapsed_time": "8h 47m 22s", "remaining_time": "2h 33m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073951} +{"loss": 0.15489502, "grad_norm": 0.52601755, "learning_rate": 1.25e-06, "tIoU": 0.56415205, "tRecall_50": 0.61875, "tRecall_75": 0.3875, "tRecall_90": 0.25625, "token_acc": 0.94414317, "epoch": 0.77687593, "global_step/max_steps": "2345/3019", "percentage": "77.67%", "elapsed_time": "8h 48m 30s", "remaining_time": "2h 31m 54s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073951} +{"loss": 0.16513672, "grad_norm": 0.67765665, "learning_rate": 1.23e-06, "tIoU": 0.51463194, "tRecall_50": 0.575, "tRecall_75": 0.4, "tRecall_90": 0.2375, "token_acc": 0.9373461, "epoch": 0.77853238, "global_step/max_steps": "2350/3019", "percentage": "77.84%", "elapsed_time": "8h 49m 37s", "remaining_time": "2h 30m 46s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073951} +{"loss": 0.16092529, "grad_norm": 0.75463802, "learning_rate": 1.22e-06, "tIoU": 0.52499647, "tRecall_50": 0.5625, "tRecall_75": 0.40625, "tRecall_90": 0.23125, "token_acc": 0.94130435, "epoch": 0.78018884, "global_step/max_steps": "2355/3019", "percentage": "78.01%", "elapsed_time": "8h 50m 43s", "remaining_time": "2h 29m 38s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073955} +{"loss": 0.1460083, "grad_norm": 0.52153689, "learning_rate": 1.2e-06, "tIoU": 0.55633059, "tRecall_50": 0.6, "tRecall_75": 0.40625, "tRecall_90": 0.3, "token_acc": 0.94576179, "epoch": 0.78184529, "global_step/max_steps": "2360/3019", "percentage": "78.17%", "elapsed_time": "8h 51m 52s", "remaining_time": "2h 28m 31s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073952} +{"loss": 0.17211914, "grad_norm": 0.79749179, "learning_rate": 1.18e-06, "tIoU": 0.51473947, "tRecall_50": 0.575, "tRecall_75": 0.34375, "tRecall_90": 0.1625, "token_acc": 0.93309763, "epoch": 0.78350174, "global_step/max_steps": "2365/3019", "percentage": "78.34%", "elapsed_time": "8h 52m 57s", "remaining_time": "2h 27m 22s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073959} +{"loss": 0.16275635, "grad_norm": 0.3969059, "learning_rate": 1.16e-06, "tIoU": 0.50973334, "tRecall_50": 0.58125, "tRecall_75": 0.35, "tRecall_90": 0.2125, "token_acc": 0.93704206, "epoch": 0.78515819, "global_step/max_steps": "2370/3019", "percentage": "78.50%", "elapsed_time": "8h 53m 59s", "remaining_time": "2h 26m 13s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07397} +{"loss": 0.15360107, "grad_norm": 0.48728055, "learning_rate": 1.15e-06, "tIoU": 0.54811215, "tRecall_50": 0.575, "tRecall_75": 0.4, "tRecall_90": 0.28125, "token_acc": 0.94230243, "epoch": 0.78681464, "global_step/max_steps": "2375/3019", "percentage": "78.67%", "elapsed_time": "8h 55m 6s", "remaining_time": "2h 25m 6s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073972} +{"loss": 0.15331421, "grad_norm": 0.64290798, "learning_rate": 1.13e-06, "tIoU": 0.55249734, "tRecall_50": 0.59375, "tRecall_75": 0.4125, "tRecall_90": 0.26875, "token_acc": 0.94274084, "epoch": 0.78847109, "global_step/max_steps": "2380/3019", "percentage": "78.83%", "elapsed_time": "8h 56m 11s", "remaining_time": "2h 23m 57s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073978} +{"loss": 0.15187988, "grad_norm": 0.45479348, "learning_rate": 1.11e-06, "tIoU": 0.53728623, "tRecall_50": 0.5875, "tRecall_75": 0.39375, "tRecall_90": 0.2375, "token_acc": 0.93941886, "epoch": 0.79012755, "global_step/max_steps": "2385/3019", "percentage": "79.00%", "elapsed_time": "8h 57m 20s", "remaining_time": "2h 22m 50s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073974} +{"loss": 0.16217041, "grad_norm": 0.45014676, "learning_rate": 1.1e-06, "tIoU": 0.56425982, "tRecall_50": 0.63125, "tRecall_75": 0.41875, "tRecall_90": 0.3125, "token_acc": 0.9434117, "epoch": 0.791784, "global_step/max_steps": "2390/3019", "percentage": "79.17%", "elapsed_time": "8h 58m 26s", "remaining_time": "2h 21m 42s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073979} +{"loss": 0.17185059, "grad_norm": 0.58433652, "learning_rate": 1.08e-06, "tIoU": 0.50322981, "tRecall_50": 0.56875, "tRecall_75": 0.3125, "tRecall_90": 0.2, "token_acc": 0.93493712, "epoch": 0.79344045, "global_step/max_steps": "2395/3019", "percentage": "79.33%", "elapsed_time": "8h 59m 32s", "remaining_time": "2h 20m 34s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073983} +{"loss": 0.15487061, "grad_norm": 0.75916237, "learning_rate": 1.06e-06, "tIoU": 0.56146755, "tRecall_50": 0.6, "tRecall_75": 0.41875, "tRecall_90": 0.3, "token_acc": 0.94573221, "epoch": 0.7950969, "global_step/max_steps": "2400/3019", "percentage": "79.50%", "elapsed_time": "9h 0m 43s", "remaining_time": "2h 19m 27s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073976} +{"loss": 0.16737061, "grad_norm": 0.6303367, "learning_rate": 1.05e-06, "tIoU": 0.49944859, "tRecall_50": 0.55625, "tRecall_75": 0.3375, "tRecall_90": 0.21875, "token_acc": 0.9360892, "epoch": 0.79675335, "global_step/max_steps": "2405/3019", "percentage": "79.66%", "elapsed_time": "9h 1m 45s", "remaining_time": "2h 18m 18s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073987} +{"loss": 0.16975098, "grad_norm": 0.83850056, "learning_rate": 1.03e-06, "tIoU": 0.55407686, "tRecall_50": 0.625, "tRecall_75": 0.38125, "tRecall_90": 0.25, "token_acc": 0.94136897, "epoch": 0.79840981, "global_step/max_steps": "2410/3019", "percentage": "79.83%", "elapsed_time": "9h 2m 59s", "remaining_time": "2h 17m 12s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073974} +{"loss": 0.14515381, "grad_norm": 0.62498134, "learning_rate": 1.01e-06, "tIoU": 0.59329076, "tRecall_50": 0.66875, "tRecall_75": 0.4625, "tRecall_90": 0.28125, "token_acc": 0.94686387, "epoch": 0.80006626, "global_step/max_steps": "2415/3019", "percentage": "79.99%", "elapsed_time": "9h 4m 5s", "remaining_time": "2h 16m 4s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073977} +{"loss": 0.16220703, "grad_norm": 2.34671497, "learning_rate": 1e-06, "tIoU": 0.47458863, "tRecall_50": 0.51875, "tRecall_75": 0.31875, "tRecall_90": 0.20625, "token_acc": 0.93780687, "epoch": 0.80172271, "global_step/max_steps": "2420/3019", "percentage": "80.16%", "elapsed_time": "9h 5m 13s", "remaining_time": "2h 14m 57s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073975} +{"loss": 0.15524902, "grad_norm": 0.69216126, "learning_rate": 9.8e-07, "tIoU": 0.53832912, "tRecall_50": 0.54375, "tRecall_75": 0.4, "tRecall_90": 0.3, "token_acc": 0.9442623, "epoch": 0.80337916, "global_step/max_steps": "2425/3019", "percentage": "80.32%", "elapsed_time": "9h 6m 20s", "remaining_time": "2h 13m 49s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073978} +{"loss": 0.15458984, "grad_norm": 0.70467418, "learning_rate": 9.7e-07, "tIoU": 0.5690701, "tRecall_50": 0.60625, "tRecall_75": 0.45625, "tRecall_90": 0.25, "token_acc": 0.94486692, "epoch": 0.80503561, "global_step/max_steps": "2430/3019", "percentage": "80.49%", "elapsed_time": "9h 7m 32s", "remaining_time": "2h 12m 43s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073966} +{"loss": 0.15469971, "grad_norm": 0.51671314, "learning_rate": 9.5e-07, "tIoU": 0.52294888, "tRecall_50": 0.5625, "tRecall_75": 0.4, "tRecall_90": 0.26875, "token_acc": 0.94292736, "epoch": 0.80669207, "global_step/max_steps": "2435/3019", "percentage": "80.66%", "elapsed_time": "9h 8m 40s", "remaining_time": "2h 11m 35s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073967} +{"loss": 0.16704712, "grad_norm": 0.47278947, "learning_rate": 9.3e-07, "tIoU": 0.52114139, "tRecall_50": 0.56875, "tRecall_75": 0.4, "tRecall_90": 0.25, "token_acc": 0.93891464, "epoch": 0.80834852, "global_step/max_steps": "2440/3019", "percentage": "80.82%", "elapsed_time": "9h 9m 48s", "remaining_time": "2h 10m 28s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073965} +{"loss": 0.15396729, "grad_norm": 0.47152713, "learning_rate": 9.2e-07, "tIoU": 0.5372179, "tRecall_50": 0.56875, "tRecall_75": 0.425, "tRecall_90": 0.26875, "token_acc": 0.94288852, "epoch": 0.81000497, "global_step/max_steps": "2445/3019", "percentage": "80.99%", "elapsed_time": "9h 10m 52s", "remaining_time": "2h 9m 19s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073972} +{"loss": 0.15570679, "grad_norm": 0.53462183, "learning_rate": 9e-07, "tIoU": 0.52619805, "tRecall_50": 0.5625, "tRecall_75": 0.40625, "tRecall_90": 0.28125, "token_acc": 0.94146474, "epoch": 0.81166142, "global_step/max_steps": "2450/3019", "percentage": "81.15%", "elapsed_time": "9h 11m 59s", "remaining_time": "2h 8m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073975} +{"loss": 0.15142822, "grad_norm": 0.51185644, "learning_rate": 8.9e-07, "tIoU": 0.55739202, "tRecall_50": 0.625, "tRecall_75": 0.4125, "tRecall_90": 0.2625, "token_acc": 0.94461037, "epoch": 0.81331787, "global_step/max_steps": "2455/3019", "percentage": "81.32%", "elapsed_time": "9h 13m 9s", "remaining_time": "2h 7m 4s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073968} +{"loss": 0.15771484, "grad_norm": 0.56790376, "learning_rate": 8.7e-07, "tIoU": 0.54014616, "tRecall_50": 0.5875, "tRecall_75": 0.425, "tRecall_90": 0.2875, "token_acc": 0.94417077, "epoch": 0.81497432, "global_step/max_steps": "2460/3019", "percentage": "81.48%", "elapsed_time": "9h 14m 12s", "remaining_time": "2h 5m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073979} +{"loss": 0.1682251, "grad_norm": 0.54755485, "learning_rate": 8.6e-07, "tIoU": 0.48176703, "tRecall_50": 0.5, "tRecall_75": 0.35, "tRecall_90": 0.225, "token_acc": 0.93608304, "epoch": 0.81663078, "global_step/max_steps": "2465/3019", "percentage": "81.65%", "elapsed_time": "9h 15m 18s", "remaining_time": "2h 4m 48s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073982} +{"loss": 0.1654541, "grad_norm": 0.49165377, "learning_rate": 8.4e-07, "tIoU": 0.53725858, "tRecall_50": 0.59375, "tRecall_75": 0.41875, "tRecall_90": 0.2625, "token_acc": 0.9378763, "epoch": 0.81828723, "global_step/max_steps": "2470/3019", "percentage": "81.82%", "elapsed_time": "9h 16m 22s", "remaining_time": "2h 3m 39s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07399} +{"loss": 0.15437012, "grad_norm": 0.65674973, "learning_rate": 8.3e-07, "tIoU": 0.51907372, "tRecall_50": 0.56875, "tRecall_75": 0.4125, "tRecall_90": 0.225, "token_acc": 0.94006568, "epoch": 0.81994368, "global_step/max_steps": "2475/3019", "percentage": "81.98%", "elapsed_time": "9h 17m 31s", "remaining_time": "2h 2m 32s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073988} +{"loss": 0.15341797, "grad_norm": 0.65429068, "learning_rate": 8.1e-07, "tIoU": 0.54075923, "tRecall_50": 0.60625, "tRecall_75": 0.38125, "tRecall_90": 0.2625, "token_acc": 0.94348894, "epoch": 0.82160013, "global_step/max_steps": "2480/3019", "percentage": "82.15%", "elapsed_time": "9h 18m 45s", "remaining_time": "2h 1m 26s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073973} +{"loss": 0.16204224, "grad_norm": 0.69178367, "learning_rate": 8e-07, "tIoU": 0.50395594, "tRecall_50": 0.53125, "tRecall_75": 0.3375, "tRecall_90": 0.2375, "token_acc": 0.94014559, "epoch": 0.82325658, "global_step/max_steps": "2485/3019", "percentage": "82.31%", "elapsed_time": "9h 19m 51s", "remaining_time": "2h 0m 18s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073977} +{"loss": 0.17220459, "grad_norm": 0.69906557, "learning_rate": 7.8e-07, "tIoU": 0.50368195, "tRecall_50": 0.5375, "tRecall_75": 0.35, "tRecall_90": 0.25625, "token_acc": 0.93788988, "epoch": 0.82491304, "global_step/max_steps": "2490/3019", "percentage": "82.48%", "elapsed_time": "9h 20m 55s", "remaining_time": "1h 59m 10s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073985} +{"loss": 0.16102905, "grad_norm": 0.62859398, "learning_rate": 7.7e-07, "tIoU": 0.51258397, "tRecall_50": 0.55, "tRecall_75": 0.33125, "tRecall_90": 0.23125, "token_acc": 0.93775594, "epoch": 0.82656949, "global_step/max_steps": "2495/3019", "percentage": "82.64%", "elapsed_time": "9h 22m 1s", "remaining_time": "1h 58m 2s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073988} +{"loss": 0.16112061, "grad_norm": 0.68151492, "learning_rate": 7.6e-07, "tIoU": 0.52466784, "tRecall_50": 0.58125, "tRecall_75": 0.375, "tRecall_90": 0.21875, "token_acc": 0.93602602, "epoch": 0.82822594, "global_step/max_steps": "2500/3019", "percentage": "82.81%", "elapsed_time": "9h 23m 6s", "remaining_time": "1h 56m 53s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073995} +{"loss": 0.1623291, "grad_norm": 0.70728588, "learning_rate": 7.4e-07, "tIoU": 0.5076376, "tRecall_50": 0.5125, "tRecall_75": 0.38125, "tRecall_90": 0.225, "token_acc": 0.93971631, "epoch": 0.82988239, "global_step/max_steps": "2505/3019", "percentage": "82.97%", "elapsed_time": "9h 24m 14s", "remaining_time": "1h 55m 46s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073993} +{"loss": 0.15618896, "grad_norm": 0.55649394, "learning_rate": 7.3e-07, "tIoU": 0.54983826, "tRecall_50": 0.6, "tRecall_75": 0.4125, "tRecall_90": 0.21875, "token_acc": 0.94187328, "epoch": 0.83153884, "global_step/max_steps": "2510/3019", "percentage": "83.14%", "elapsed_time": "9h 25m 21s", "remaining_time": "1h 54m 38s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073995} +{"loss": 0.15855713, "grad_norm": 0.54246819, "learning_rate": 7.1e-07, "tIoU": 0.53365216, "tRecall_50": 0.58125, "tRecall_75": 0.3875, "tRecall_90": 0.225, "token_acc": 0.94392779, "epoch": 0.8331953, "global_step/max_steps": "2515/3019", "percentage": "83.31%", "elapsed_time": "9h 26m 30s", "remaining_time": "1h 53m 31s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073992} +{"loss": 0.16523437, "grad_norm": 0.48574844, "learning_rate": 7e-07, "tIoU": 0.52912474, "tRecall_50": 0.55, "tRecall_75": 0.40625, "tRecall_90": 0.2875, "token_acc": 0.94016393, "epoch": 0.83485175, "global_step/max_steps": "2520/3019", "percentage": "83.47%", "elapsed_time": "9h 27m 34s", "remaining_time": "1h 52m 23s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073999} +{"loss": 0.15307617, "grad_norm": 0.65652519, "learning_rate": 6.9e-07, "tIoU": 0.58006215, "tRecall_50": 0.6375, "tRecall_75": 0.45, "tRecall_90": 0.26875, "token_acc": 0.94244408, "epoch": 0.8365082, "global_step/max_steps": "2525/3019", "percentage": "83.64%", "elapsed_time": "9h 28m 48s", "remaining_time": "1h 51m 17s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073985} +{"loss": 0.150354, "grad_norm": 0.73899788, "learning_rate": 6.7e-07, "tIoU": 0.56403179, "tRecall_50": 0.6125, "tRecall_75": 0.425, "tRecall_90": 0.2625, "token_acc": 0.94003286, "epoch": 0.83816465, "global_step/max_steps": "2530/3019", "percentage": "83.80%", "elapsed_time": "9h 29m 54s", "remaining_time": "1h 50m 9s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07399} +{"loss": 0.16765137, "grad_norm": 0.72637975, "learning_rate": 6.6e-07, "tIoU": 0.49884806, "tRecall_50": 0.53125, "tRecall_75": 0.34375, "tRecall_90": 0.2375, "token_acc": 0.93858696, "epoch": 0.8398211, "global_step/max_steps": "2535/3019", "percentage": "83.97%", "elapsed_time": "9h 31m 4s", "remaining_time": "1h 49m 1s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073984} +{"loss": 0.16037598, "grad_norm": 0.68379599, "learning_rate": 6.5e-07, "tIoU": 0.51441016, "tRecall_50": 0.50625, "tRecall_75": 0.3875, "tRecall_90": 0.2875, "token_acc": 0.93934426, "epoch": 0.84147756, "global_step/max_steps": "2540/3019", "percentage": "84.13%", "elapsed_time": "9h 32m 14s", "remaining_time": "1h 47m 54s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073979} +{"loss": 0.15837402, "grad_norm": 0.55650425, "learning_rate": 6.3e-07, "tIoU": 0.53583217, "tRecall_50": 0.56875, "tRecall_75": 0.375, "tRecall_90": 0.23125, "token_acc": 0.940586, "epoch": 0.84313401, "global_step/max_steps": "2545/3019", "percentage": "84.30%", "elapsed_time": "9h 33m 19s", "remaining_time": "1h 46m 46s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073984} +{"loss": 0.17939453, "grad_norm": 0.49884149, "learning_rate": 6.2e-07, "tIoU": 0.53002648, "tRecall_50": 0.5375, "tRecall_75": 0.40625, "tRecall_90": 0.28125, "token_acc": 0.93787466, "epoch": 0.84479046, "global_step/max_steps": "2550/3019", "percentage": "84.47%", "elapsed_time": "9h 34m 27s", "remaining_time": "1h 45m 39s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073983} +{"loss": 0.16768799, "grad_norm": 0.65434706, "learning_rate": 6.1e-07, "tIoU": 0.50537491, "tRecall_50": 0.51875, "tRecall_75": 0.3625, "tRecall_90": 0.28125, "token_acc": 0.93435031, "epoch": 0.84644691, "global_step/max_steps": "2555/3019", "percentage": "84.63%", "elapsed_time": "9h 35m 32s", "remaining_time": "1h 44m 31s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073988} +{"loss": 0.16727295, "grad_norm": 0.97917861, "learning_rate": 5.9e-07, "tIoU": 0.5062737, "tRecall_50": 0.55625, "tRecall_75": 0.29375, "tRecall_90": 0.2125, "token_acc": 0.93736324, "epoch": 0.84810336, "global_step/max_steps": "2560/3019", "percentage": "84.80%", "elapsed_time": "9h 36m 40s", "remaining_time": "1h 43m 23s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073987} +{"loss": 0.16414185, "grad_norm": 0.68624604, "learning_rate": 5.8e-07, "tIoU": 0.50867667, "tRecall_50": 0.54375, "tRecall_75": 0.375, "tRecall_90": 0.25625, "token_acc": 0.93974918, "epoch": 0.84975981, "global_step/max_steps": "2565/3019", "percentage": "84.96%", "elapsed_time": "9h 37m 49s", "remaining_time": "1h 42m 16s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073985} +{"loss": 0.1460083, "grad_norm": 0.52615255, "learning_rate": 5.7e-07, "tIoU": 0.56607313, "tRecall_50": 0.6, "tRecall_75": 0.41875, "tRecall_90": 0.2625, "token_acc": 0.94551893, "epoch": 0.85141627, "global_step/max_steps": "2570/3019", "percentage": "85.13%", "elapsed_time": "9h 38m 54s", "remaining_time": "1h 41m 8s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07399} +{"loss": 0.16983643, "grad_norm": 0.68742061, "learning_rate": 5.6e-07, "tIoU": 0.53418131, "tRecall_50": 0.5625, "tRecall_75": 0.40625, "tRecall_90": 0.2875, "token_acc": 0.93673966, "epoch": 0.85307272, "global_step/max_steps": "2575/3019", "percentage": "85.29%", "elapsed_time": "9h 39m 57s", "remaining_time": "1h 40m 0s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073999} +{"loss": 0.16873779, "grad_norm": 0.55257684, "learning_rate": 5.4e-07, "tIoU": 0.49448662, "tRecall_50": 0.5625, "tRecall_75": 0.33125, "tRecall_90": 0.16875, "token_acc": 0.9372442, "epoch": 0.85472917, "global_step/max_steps": "2580/3019", "percentage": "85.46%", "elapsed_time": "9h 41m 3s", "remaining_time": "1h 38m 52s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074003} +{"loss": 0.17290039, "grad_norm": 0.55129206, "learning_rate": 5.3e-07, "tIoU": 0.50668283, "tRecall_50": 0.55625, "tRecall_75": 0.35625, "tRecall_90": 0.225, "token_acc": 0.93727843, "epoch": 0.85638562, "global_step/max_steps": "2585/3019", "percentage": "85.62%", "elapsed_time": "9h 42m 9s", "remaining_time": "1h 37m 44s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074007} +{"loss": 0.17460938, "grad_norm": 1.39201772, "learning_rate": 5.2e-07, "tIoU": 0.5285748, "tRecall_50": 0.56875, "tRecall_75": 0.3875, "tRecall_90": 0.26875, "token_acc": 0.93738037, "epoch": 0.85804207, "global_step/max_steps": "2590/3019", "percentage": "85.79%", "elapsed_time": "9h 43m 12s", "remaining_time": "1h 36m 36s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074016} +{"loss": 0.17049561, "grad_norm": 0.62874097, "learning_rate": 5.1e-07, "tIoU": 0.485151, "tRecall_50": 0.51875, "tRecall_75": 0.33125, "tRecall_90": 0.1625, "token_acc": 0.93158899, "epoch": 0.85969853, "global_step/max_steps": "2595/3019", "percentage": "85.96%", "elapsed_time": "9h 44m 19s", "remaining_time": "1h 35m 28s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074018} +{"loss": 0.15293579, "grad_norm": 0.58158988, "learning_rate": 5e-07, "tIoU": 0.5580605, "tRecall_50": 0.625, "tRecall_75": 0.39375, "tRecall_90": 0.24375, "token_acc": 0.94162784, "epoch": 0.86135498, "global_step/max_steps": "2600/3019", "percentage": "86.12%", "elapsed_time": "9h 45m 23s", "remaining_time": "1h 34m 20s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074025} +{"loss": 0.15257568, "grad_norm": 0.66748154, "learning_rate": 4.9e-07, "tIoU": 0.55859623, "tRecall_50": 0.61875, "tRecall_75": 0.375, "tRecall_90": 0.2375, "token_acc": 0.94193021, "epoch": 0.86301143, "global_step/max_steps": "2605/3019", "percentage": "86.29%", "elapsed_time": "9h 46m 30s", "remaining_time": "1h 33m 12s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074026} +{"loss": 0.15527344, "grad_norm": 0.55767244, "learning_rate": 4.7e-07, "tIoU": 0.55069637, "tRecall_50": 0.60625, "tRecall_75": 0.4125, "tRecall_90": 0.28125, "token_acc": 0.93737984, "epoch": 0.86466788, "global_step/max_steps": "2610/3019", "percentage": "86.45%", "elapsed_time": "9h 47m 36s", "remaining_time": "1h 32m 4s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074029} +{"loss": 0.17331543, "grad_norm": 0.45226309, "learning_rate": 4.6e-07, "tIoU": 0.48772349, "tRecall_50": 0.53125, "tRecall_75": 0.35, "tRecall_90": 0.175, "token_acc": 0.9375338, "epoch": 0.86632433, "global_step/max_steps": "2615/3019", "percentage": "86.62%", "elapsed_time": "9h 48m 44s", "remaining_time": "1h 30m 57s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074028} +{"loss": 0.15415039, "grad_norm": 0.75766814, "learning_rate": 4.5e-07, "tIoU": 0.54727311, "tRecall_50": 0.575, "tRecall_75": 0.41875, "tRecall_90": 0.28125, "token_acc": 0.94411286, "epoch": 0.86798079, "global_step/max_steps": "2620/3019", "percentage": "86.78%", "elapsed_time": "9h 49m 50s", "remaining_time": "1h 29m 49s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07403} +{"loss": 0.15056152, "grad_norm": 0.61049849, "learning_rate": 4.4e-07, "tIoU": 0.5675601, "tRecall_50": 0.56875, "tRecall_75": 0.45625, "tRecall_90": 0.30625, "token_acc": 0.94409599, "epoch": 0.86963724, "global_step/max_steps": "2625/3019", "percentage": "86.95%", "elapsed_time": "9h 50m 53s", "remaining_time": "1h 28m 41s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07404} +{"loss": 0.15528564, "grad_norm": 0.6254884, "learning_rate": 4.3e-07, "tIoU": 0.5098196, "tRecall_50": 0.5375, "tRecall_75": 0.40625, "tRecall_90": 0.225, "token_acc": 0.9385034, "epoch": 0.87129369, "global_step/max_steps": "2630/3019", "percentage": "87.11%", "elapsed_time": "9h 51m 59s", "remaining_time": "1h 27m 33s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074044} +{"loss": 0.1647644, "grad_norm": 0.81283766, "learning_rate": 4.2e-07, "tIoU": 0.51389542, "tRecall_50": 0.5625, "tRecall_75": 0.4, "tRecall_90": 0.20625, "token_acc": 0.93659864, "epoch": 0.87295014, "global_step/max_steps": "2635/3019", "percentage": "87.28%", "elapsed_time": "9h 53m 10s", "remaining_time": "1h 26m 26s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074037} +{"loss": 0.14893188, "grad_norm": 0.5039224, "learning_rate": 4.1e-07, "tIoU": 0.51364779, "tRecall_50": 0.53125, "tRecall_75": 0.3625, "tRecall_90": 0.25, "token_acc": 0.94149512, "epoch": 0.87460659, "global_step/max_steps": "2640/3019", "percentage": "87.45%", "elapsed_time": "9h 54m 22s", "remaining_time": "1h 25m 19s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074027} +{"loss": 0.17333984, "grad_norm": 0.48948956, "learning_rate": 4e-07, "tIoU": 0.4695116, "tRecall_50": 0.4625, "tRecall_75": 0.3125, "tRecall_90": 0.18125, "token_acc": 0.93654684, "epoch": 0.87626304, "global_step/max_steps": "2645/3019", "percentage": "87.61%", "elapsed_time": "9h 55m 32s", "remaining_time": "1h 24m 12s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074021} +{"loss": 0.16063232, "grad_norm": 0.40938482, "learning_rate": 3.9e-07, "tIoU": 0.49820685, "tRecall_50": 0.5375, "tRecall_75": 0.35625, "tRecall_90": 0.24375, "token_acc": 0.93705528, "epoch": 0.8779195, "global_step/max_steps": "2650/3019", "percentage": "87.78%", "elapsed_time": "9h 56m 37s", "remaining_time": "1h 23m 4s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074027} +{"loss": 0.16600952, "grad_norm": 0.52045435, "learning_rate": 3.8e-07, "tIoU": 0.53276134, "tRecall_50": 0.5625, "tRecall_75": 0.3875, "tRecall_90": 0.24375, "token_acc": 0.93998909, "epoch": 0.87957595, "global_step/max_steps": "2655/3019", "percentage": "87.94%", "elapsed_time": "9h 57m 42s", "remaining_time": "1h 21m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074032} +{"loss": 0.15164795, "grad_norm": 0.69445091, "learning_rate": 3.7e-07, "tIoU": 0.54920058, "tRecall_50": 0.60625, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.94180038, "epoch": 0.8812324, "global_step/max_steps": "2660/3019", "percentage": "88.11%", "elapsed_time": "9h 58m 49s", "remaining_time": "1h 20m 49s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074035} +{"loss": 0.14561157, "grad_norm": 0.7271902, "learning_rate": 3.6e-07, "tIoU": 0.55342016, "tRecall_50": 0.6125, "tRecall_75": 0.425, "tRecall_90": 0.30625, "token_acc": 0.94396083, "epoch": 0.88288885, "global_step/max_steps": "2665/3019", "percentage": "88.27%", "elapsed_time": "10h 0m 0s", "remaining_time": "1h 19m 42s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074026} +{"loss": 0.15371094, "grad_norm": 0.61312395, "learning_rate": 3.5e-07, "tIoU": 0.53571792, "tRecall_50": 0.575, "tRecall_75": 0.4125, "tRecall_90": 0.26875, "token_acc": 0.94261626, "epoch": 0.8845453, "global_step/max_steps": "2670/3019", "percentage": "88.44%", "elapsed_time": "10h 1m 5s", "remaining_time": "1h 18m 34s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074032} +{"loss": 0.15107422, "grad_norm": 0.49681997, "learning_rate": 3.4e-07, "tIoU": 0.57655559, "tRecall_50": 0.63125, "tRecall_75": 0.43125, "tRecall_90": 0.25625, "token_acc": 0.94291927, "epoch": 0.88620176, "global_step/max_steps": "2675/3019", "percentage": "88.61%", "elapsed_time": "10h 2m 11s", "remaining_time": "1h 17m 26s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074035} +{"loss": 0.15844727, "grad_norm": 0.50766146, "learning_rate": 3.3e-07, "tIoU": 0.53414268, "tRecall_50": 0.55, "tRecall_75": 0.375, "tRecall_90": 0.25625, "token_acc": 0.94071254, "epoch": 0.88785821, "global_step/max_steps": "2680/3019", "percentage": "88.77%", "elapsed_time": "10h 3m 16s", "remaining_time": "1h 16m 18s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07404} +{"loss": 0.14836426, "grad_norm": 0.61267465, "learning_rate": 3.2e-07, "tIoU": 0.57533302, "tRecall_50": 0.65, "tRecall_75": 0.45625, "tRecall_90": 0.3, "token_acc": 0.94124078, "epoch": 0.88951466, "global_step/max_steps": "2685/3019", "percentage": "88.94%", "elapsed_time": "10h 4m 26s", "remaining_time": "1h 15m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074035} +{"loss": 0.17729492, "grad_norm": 0.75557655, "learning_rate": 3.1e-07, "tIoU": 0.46550651, "tRecall_50": 0.48125, "tRecall_75": 0.34375, "tRecall_90": 0.2125, "token_acc": 0.93037803, "epoch": 0.89117111, "global_step/max_steps": "2690/3019", "percentage": "89.10%", "elapsed_time": "10h 5m 32s", "remaining_time": "1h 14m 3s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074039} +{"loss": 0.14302368, "grad_norm": 0.76927644, "learning_rate": 3e-07, "tIoU": 0.56302214, "tRecall_50": 0.625, "tRecall_75": 0.425, "tRecall_90": 0.3, "token_acc": 0.94450519, "epoch": 0.89282756, "global_step/max_steps": "2695/3019", "percentage": "89.27%", "elapsed_time": "10h 6m 40s", "remaining_time": "1h 12m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074037} +{"loss": 0.14648438, "grad_norm": 0.57204282, "learning_rate": 2.9e-07, "tIoU": 0.54900799, "tRecall_50": 0.6, "tRecall_75": 0.4125, "tRecall_90": 0.275, "token_acc": 0.94263862, "epoch": 0.89448402, "global_step/max_steps": "2700/3019", "percentage": "89.43%", "elapsed_time": "10h 7m 47s", "remaining_time": "1h 11m 48s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074038} +{"loss": 0.15216675, "grad_norm": 0.86713183, "learning_rate": 2.8e-07, "tIoU": 0.57141962, "tRecall_50": 0.63125, "tRecall_75": 0.41875, "tRecall_90": 0.23125, "token_acc": 0.9429042, "epoch": 0.89614047, "global_step/max_steps": "2705/3019", "percentage": "89.60%", "elapsed_time": "10h 8m 54s", "remaining_time": "1h 10m 41s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074039} +{"loss": 0.14770508, "grad_norm": 0.63123131, "learning_rate": 2.7e-07, "tIoU": 0.56925473, "tRecall_50": 0.61875, "tRecall_75": 0.43125, "tRecall_90": 0.26875, "token_acc": 0.94523161, "epoch": 0.89779692, "global_step/max_steps": "2710/3019", "percentage": "89.76%", "elapsed_time": "10h 10m 1s", "remaining_time": "1h 9m 33s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074041} +{"loss": 0.15697021, "grad_norm": 0.96746069, "learning_rate": 2.6e-07, "tIoU": 0.5186336, "tRecall_50": 0.575, "tRecall_75": 0.375, "tRecall_90": 0.23125, "token_acc": 0.94013605, "epoch": 0.89945337, "global_step/max_steps": "2715/3019", "percentage": "89.93%", "elapsed_time": "10h 11m 10s", "remaining_time": "1h 8m 26s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074037} +{"loss": 0.15101318, "grad_norm": 0.53268993, "learning_rate": 2.6e-07, "tIoU": 0.56083002, "tRecall_50": 0.5875, "tRecall_75": 0.45625, "tRecall_90": 0.3, "token_acc": 0.9469075, "epoch": 0.90110982, "global_step/max_steps": "2720/3019", "percentage": "90.10%", "elapsed_time": "10h 12m 20s", "remaining_time": "1h 7m 18s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074033} +{"loss": 0.16943359, "grad_norm": 0.51062196, "learning_rate": 2.5e-07, "tIoU": 0.48913627, "tRecall_50": 0.53125, "tRecall_75": 0.36875, "tRecall_90": 0.19375, "token_acc": 0.93320663, "epoch": 0.90276627, "global_step/max_steps": "2725/3019", "percentage": "90.26%", "elapsed_time": "10h 13m 35s", "remaining_time": "1h 6m 12s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074018} +{"loss": 0.16939087, "grad_norm": 0.62513632, "learning_rate": 2.4e-07, "tIoU": 0.48248038, "tRecall_50": 0.53125, "tRecall_75": 0.31875, "tRecall_90": 0.19375, "token_acc": 0.93440831, "epoch": 0.90442273, "global_step/max_steps": "2730/3019", "percentage": "90.43%", "elapsed_time": "10h 14m 40s", "remaining_time": "1h 5m 4s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074022} +{"loss": 0.14824219, "grad_norm": 0.61283821, "learning_rate": 2.3e-07, "tIoU": 0.54461689, "tRecall_50": 0.575, "tRecall_75": 0.41875, "tRecall_90": 0.2875, "token_acc": 0.94408252, "epoch": 0.90607918, "global_step/max_steps": "2735/3019", "percentage": "90.59%", "elapsed_time": "10h 15m 46s", "remaining_time": "1h 3m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074026} +{"loss": 0.15986938, "grad_norm": 1.07309163, "learning_rate": 2.2e-07, "tIoU": 0.49481511, "tRecall_50": 0.5375, "tRecall_75": 0.3375, "tRecall_90": 0.25, "token_acc": 0.94088937, "epoch": 0.90773563, "global_step/max_steps": "2740/3019", "percentage": "90.76%", "elapsed_time": "10h 16m 53s", "remaining_time": "1h 2m 48s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074028} +{"loss": 0.15233154, "grad_norm": 0.56090045, "learning_rate": 2.1e-07, "tIoU": 0.56277933, "tRecall_50": 0.60625, "tRecall_75": 0.425, "tRecall_90": 0.325, "token_acc": 0.9446414, "epoch": 0.90939208, "global_step/max_steps": "2745/3019", "percentage": "90.92%", "elapsed_time": "10h 18m 0s", "remaining_time": "1h 1m 41s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074028} +{"loss": 0.15305786, "grad_norm": 0.65905458, "learning_rate": 2.1e-07, "tIoU": 0.54942904, "tRecall_50": 0.63125, "tRecall_75": 0.35625, "tRecall_90": 0.275, "token_acc": 0.94170526, "epoch": 0.91104853, "global_step/max_steps": "2750/3019", "percentage": "91.09%", "elapsed_time": "10h 19m 12s", "remaining_time": "1h 0m 34s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07402} +{"loss": 0.17062988, "grad_norm": 0.45768371, "learning_rate": 2e-07, "tIoU": 0.52331986, "tRecall_50": 0.55, "tRecall_75": 0.39375, "tRecall_90": 0.23125, "token_acc": 0.93464052, "epoch": 0.91270499, "global_step/max_steps": "2755/3019", "percentage": "91.26%", "elapsed_time": "10h 20m 25s", "remaining_time": "59m 27s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074009} +{"loss": 0.15598145, "grad_norm": 0.69677293, "learning_rate": 1.9e-07, "tIoU": 0.55192471, "tRecall_50": 0.6, "tRecall_75": 0.4125, "tRecall_90": 0.24375, "token_acc": 0.93837838, "epoch": 0.91436144, "global_step/max_steps": "2760/3019", "percentage": "91.42%", "elapsed_time": "10h 21m 36s", "remaining_time": "58m 19s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.074002} +{"loss": 0.17286377, "grad_norm": 0.44120714, "learning_rate": 1.8e-07, "tIoU": 0.477634, "tRecall_50": 0.5125, "tRecall_75": 0.34375, "tRecall_90": 0.225, "token_acc": 0.93329718, "epoch": 0.91601789, "global_step/max_steps": "2765/3019", "percentage": "91.59%", "elapsed_time": "10h 22m 46s", "remaining_time": "57m 12s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073998} +{"loss": 0.1543335, "grad_norm": 0.46227229, "learning_rate": 1.8e-07, "tIoU": 0.53533635, "tRecall_50": 0.575, "tRecall_75": 0.4, "tRecall_90": 0.24375, "token_acc": 0.93833198, "epoch": 0.91767434, "global_step/max_steps": "2770/3019", "percentage": "91.75%", "elapsed_time": "10h 23m 59s", "remaining_time": "56m 5s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073986} +{"loss": 0.15783691, "grad_norm": 0.71051216, "learning_rate": 1.7e-07, "tIoU": 0.5579864, "tRecall_50": 0.625, "tRecall_75": 0.425, "tRecall_90": 0.275, "token_acc": 0.94095134, "epoch": 0.91933079, "global_step/max_steps": "2775/3019", "percentage": "91.92%", "elapsed_time": "10h 25m 8s", "remaining_time": "54m 58s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073983} +{"loss": 0.18677979, "grad_norm": 0.51626295, "learning_rate": 1.6e-07, "tIoU": 0.46270094, "tRecall_50": 0.48125, "tRecall_75": 0.35625, "tRecall_90": 0.2375, "token_acc": 0.93, "epoch": 0.92098725, "global_step/max_steps": "2780/3019", "percentage": "92.08%", "elapsed_time": "10h 26m 13s", "remaining_time": "53m 50s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073988} +{"loss": 0.15511475, "grad_norm": 0.56346881, "learning_rate": 1.6e-07, "tIoU": 0.54569769, "tRecall_50": 0.6, "tRecall_75": 0.44375, "tRecall_90": 0.2625, "token_acc": 0.93931184, "epoch": 0.9226437, "global_step/max_steps": "2785/3019", "percentage": "92.25%", "elapsed_time": "10h 27m 21s", "remaining_time": "52m 42s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073987} +{"loss": 0.16069946, "grad_norm": 0.64886898, "learning_rate": 1.5e-07, "tIoU": 0.55492249, "tRecall_50": 0.6, "tRecall_75": 0.41875, "tRecall_90": 0.2625, "token_acc": 0.94212394, "epoch": 0.92430015, "global_step/max_steps": "2790/3019", "percentage": "92.41%", "elapsed_time": "10h 28m 33s", "remaining_time": "51m 35s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07398} +{"loss": 0.15605469, "grad_norm": 0.5581038, "learning_rate": 1.4e-07, "tIoU": 0.5302978, "tRecall_50": 0.5875, "tRecall_75": 0.3875, "tRecall_90": 0.25, "token_acc": 0.93797814, "epoch": 0.9259566, "global_step/max_steps": "2795/3019", "percentage": "92.58%", "elapsed_time": "10h 29m 46s", "remaining_time": "50m 28s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073969} +{"loss": 0.1664917, "grad_norm": 1.2052716, "learning_rate": 1.4e-07, "tIoU": 0.55261168, "tRecall_50": 0.60625, "tRecall_75": 0.4, "tRecall_90": 0.24375, "token_acc": 0.94171862, "epoch": 0.92761305, "global_step/max_steps": "2800/3019", "percentage": "92.75%", "elapsed_time": "10h 30m 58s", "remaining_time": "49m 21s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073959} +{"loss": 0.14595947, "grad_norm": 0.58044571, "learning_rate": 1.3e-07, "tIoU": 0.57653322, "tRecall_50": 0.61875, "tRecall_75": 0.4375, "tRecall_90": 0.325, "token_acc": 0.94309828, "epoch": 0.9292695, "global_step/max_steps": "2805/3019", "percentage": "92.91%", "elapsed_time": "10h 32m 0s", "remaining_time": "48m 13s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07397} +{"loss": 0.16298828, "grad_norm": 0.57867974, "learning_rate": 1.3e-07, "tIoU": 0.52110462, "tRecall_50": 0.5875, "tRecall_75": 0.35, "tRecall_90": 0.23125, "token_acc": 0.93439957, "epoch": 0.93092596, "global_step/max_steps": "2810/3019", "percentage": "93.08%", "elapsed_time": "10h 33m 9s", "remaining_time": "47m 5s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073968} +{"loss": 0.15180664, "grad_norm": 0.73171997, "learning_rate": 1.2e-07, "tIoU": 0.51161261, "tRecall_50": 0.54375, "tRecall_75": 0.38125, "tRecall_90": 0.28125, "token_acc": 0.93892084, "epoch": 0.93258241, "global_step/max_steps": "2815/3019", "percentage": "93.24%", "elapsed_time": "10h 34m 15s", "remaining_time": "45m 57s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073971} +{"loss": 0.15935059, "grad_norm": 0.52198738, "learning_rate": 1.1e-07, "tIoU": 0.55949287, "tRecall_50": 0.5875, "tRecall_75": 0.41875, "tRecall_90": 0.3125, "token_acc": 0.94412802, "epoch": 0.93423886, "global_step/max_steps": "2820/3019", "percentage": "93.41%", "elapsed_time": "10h 35m 26s", "remaining_time": "44m 50s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073965} +{"loss": 0.16157227, "grad_norm": 0.65679353, "learning_rate": 1.1e-07, "tIoU": 0.52461605, "tRecall_50": 0.6, "tRecall_75": 0.3875, "tRecall_90": 0.225, "token_acc": 0.93887061, "epoch": 0.93589531, "global_step/max_steps": "2825/3019", "percentage": "93.57%", "elapsed_time": "10h 36m 31s", "remaining_time": "43m 42s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07397} +{"loss": 0.15240479, "grad_norm": 0.68120283, "learning_rate": 1e-07, "tIoU": 0.56289415, "tRecall_50": 0.58125, "tRecall_75": 0.425, "tRecall_90": 0.29375, "token_acc": 0.94239694, "epoch": 0.93755176, "global_step/max_steps": "2830/3019", "percentage": "93.74%", "elapsed_time": "10h 37m 44s", "remaining_time": "42m 35s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07396} +{"loss": 0.16417236, "grad_norm": 0.49755436, "learning_rate": 1e-07, "tIoU": 0.52630601, "tRecall_50": 0.58125, "tRecall_75": 0.375, "tRecall_90": 0.23125, "token_acc": 0.94153005, "epoch": 0.93920822, "global_step/max_steps": "2835/3019", "percentage": "93.91%", "elapsed_time": "10h 38m 50s", "remaining_time": "41m 27s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073962} +{"loss": 0.15678711, "grad_norm": 0.61638951, "learning_rate": 9e-08, "tIoU": 0.54905934, "tRecall_50": 0.5875, "tRecall_75": 0.4125, "tRecall_90": 0.2875, "token_acc": 0.94037571, "epoch": 0.94086467, "global_step/max_steps": "2840/3019", "percentage": "94.07%", "elapsed_time": "10h 39m 56s", "remaining_time": "40m 20s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073965} +{"loss": 0.16193848, "grad_norm": 0.5530082, "learning_rate": 9e-08, "tIoU": 0.50227232, "tRecall_50": 0.5375, "tRecall_75": 0.36875, "tRecall_90": 0.18125, "token_acc": 0.93643032, "epoch": 0.94252112, "global_step/max_steps": "2845/3019", "percentage": "94.24%", "elapsed_time": "10h 41m 4s", "remaining_time": "39m 12s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073964} +{"loss": 0.15866699, "grad_norm": 0.67256337, "learning_rate": 8e-08, "tIoU": 0.55017467, "tRecall_50": 0.61875, "tRecall_75": 0.41875, "tRecall_90": 0.18125, "token_acc": 0.93955894, "epoch": 0.94417757, "global_step/max_steps": "2850/3019", "percentage": "94.40%", "elapsed_time": "10h 42m 11s", "remaining_time": "38m 4s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073965} +{"loss": 0.14420166, "grad_norm": 0.45273101, "learning_rate": 8e-08, "tIoU": 0.53565435, "tRecall_50": 0.54375, "tRecall_75": 0.44375, "tRecall_90": 0.275, "token_acc": 0.94384156, "epoch": 0.94583402, "global_step/max_steps": "2855/3019", "percentage": "94.57%", "elapsed_time": "10h 43m 19s", "remaining_time": "36m 57s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073965} +{"loss": 0.15944824, "grad_norm": 0.57842362, "learning_rate": 7e-08, "tIoU": 0.54485531, "tRecall_50": 0.58125, "tRecall_75": 0.4625, "tRecall_90": 0.3, "token_acc": 0.94016393, "epoch": 0.94749048, "global_step/max_steps": "2860/3019", "percentage": "94.73%", "elapsed_time": "10h 44m 25s", "remaining_time": "35m 49s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073967} +{"loss": 0.159729, "grad_norm": 0.60653317, "learning_rate": 7e-08, "tIoU": 0.49558449, "tRecall_50": 0.5, "tRecall_75": 0.36875, "tRecall_90": 0.21875, "token_acc": 0.93420693, "epoch": 0.94914693, "global_step/max_steps": "2865/3019", "percentage": "94.90%", "elapsed_time": "10h 45m 31s", "remaining_time": "34m 41s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073971} +{"loss": 0.15238037, "grad_norm": 0.5477348, "learning_rate": 6e-08, "tIoU": 0.55787201, "tRecall_50": 0.59375, "tRecall_75": 0.425, "tRecall_90": 0.25625, "token_acc": 0.94231821, "epoch": 0.95080338, "global_step/max_steps": "2870/3019", "percentage": "95.06%", "elapsed_time": "10h 46m 38s", "remaining_time": "33m 34s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073972} +{"loss": 0.15859375, "grad_norm": 0.53369868, "learning_rate": 6e-08, "tIoU": 0.53139524, "tRecall_50": 0.575, "tRecall_75": 0.38125, "tRecall_90": 0.275, "token_acc": 0.94242838, "epoch": 0.95245983, "global_step/max_steps": "2875/3019", "percentage": "95.23%", "elapsed_time": "10h 47m 41s", "remaining_time": "32m 26s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07398} +{"loss": 0.14896851, "grad_norm": 0.5611313, "learning_rate": 6e-08, "tIoU": 0.56537981, "tRecall_50": 0.61875, "tRecall_75": 0.46875, "tRecall_90": 0.3, "token_acc": 0.94548923, "epoch": 0.95411628, "global_step/max_steps": "2880/3019", "percentage": "95.40%", "elapsed_time": "10h 48m 53s", "remaining_time": "31m 19s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073973} +{"loss": 0.14414673, "grad_norm": 0.33091646, "learning_rate": 5e-08, "tIoU": 0.60373178, "tRecall_50": 0.675, "tRecall_75": 0.4625, "tRecall_90": 0.29375, "token_acc": 0.94703622, "epoch": 0.95577273, "global_step/max_steps": "2885/3019", "percentage": "95.56%", "elapsed_time": "10h 50m 1s", "remaining_time": "30m 11s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073971} +{"loss": 0.15085449, "grad_norm": 0.63865596, "learning_rate": 5e-08, "tIoU": 0.51421537, "tRecall_50": 0.60625, "tRecall_75": 0.3375, "tRecall_90": 0.2, "token_acc": 0.93719279, "epoch": 0.95742919, "global_step/max_steps": "2890/3019", "percentage": "95.73%", "elapsed_time": "10h 51m 11s", "remaining_time": "29m 4s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073967} +{"loss": 0.16079102, "grad_norm": 0.74770612, "learning_rate": 4e-08, "tIoU": 0.52515349, "tRecall_50": 0.575, "tRecall_75": 0.3625, "tRecall_90": 0.21875, "token_acc": 0.93715847, "epoch": 0.95908564, "global_step/max_steps": "2895/3019", "percentage": "95.89%", "elapsed_time": "10h 52m 17s", "remaining_time": "27m 56s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073971} +{"loss": 0.1550415, "grad_norm": 0.42086467, "learning_rate": 4e-08, "tIoU": 0.57164917, "tRecall_50": 0.625, "tRecall_75": 0.43125, "tRecall_90": 0.29375, "token_acc": 0.94342717, "epoch": 0.96074209, "global_step/max_steps": "2900/3019", "percentage": "96.06%", "elapsed_time": "10h 53m 29s", "remaining_time": "26m 48s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073962} +{"loss": 0.15055542, "grad_norm": 0.79154599, "learning_rate": 4e-08, "tIoU": 0.55311522, "tRecall_50": 0.6, "tRecall_75": 0.40625, "tRecall_90": 0.2875, "token_acc": 0.94452036, "epoch": 0.96239854, "global_step/max_steps": "2905/3019", "percentage": "96.22%", "elapsed_time": "10h 54m 36s", "remaining_time": "25m 41s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073962} +{"loss": 0.15845337, "grad_norm": 0.61005408, "learning_rate": 3e-08, "tIoU": 0.57693176, "tRecall_50": 0.6375, "tRecall_75": 0.45, "tRecall_90": 0.28125, "token_acc": 0.94024557, "epoch": 0.96405499, "global_step/max_steps": "2910/3019", "percentage": "96.39%", "elapsed_time": "10h 55m 42s", "remaining_time": "24m 33s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073965} +{"loss": 0.16090088, "grad_norm": 0.5148958, "learning_rate": 3e-08, "tIoU": 0.52562289, "tRecall_50": 0.6, "tRecall_75": 0.39375, "tRecall_90": 0.225, "token_acc": 0.93838219, "epoch": 0.96571145, "global_step/max_steps": "2915/3019", "percentage": "96.56%", "elapsed_time": "10h 56m 45s", "remaining_time": "23m 25s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073974} +{"loss": 0.15223389, "grad_norm": 0.46044782, "learning_rate": 3e-08, "tIoU": 0.52960515, "tRecall_50": 0.58125, "tRecall_75": 0.39375, "tRecall_90": 0.23125, "token_acc": 0.93997827, "epoch": 0.9673679, "global_step/max_steps": "2920/3019", "percentage": "96.72%", "elapsed_time": "10h 57m 51s", "remaining_time": "22m 18s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073977} +{"loss": 0.17304687, "grad_norm": 0.51511294, "learning_rate": 3e-08, "tIoU": 0.47686148, "tRecall_50": 0.5, "tRecall_75": 0.3375, "tRecall_90": 0.2125, "token_acc": 0.93816689, "epoch": 0.96902435, "global_step/max_steps": "2925/3019", "percentage": "96.89%", "elapsed_time": "10h 58m 54s", "remaining_time": "21m 10s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073985} +{"loss": 0.16057129, "grad_norm": 0.65865785, "learning_rate": 2e-08, "tIoU": 0.54304686, "tRecall_50": 0.59375, "tRecall_75": 0.41875, "tRecall_90": 0.25625, "token_acc": 0.93907104, "epoch": 0.9706808, "global_step/max_steps": "2930/3019", "percentage": "97.05%", "elapsed_time": "11h 0m 4s", "remaining_time": "20m 3s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073981} +{"loss": 0.15913086, "grad_norm": 0.75296897, "learning_rate": 2e-08, "tIoU": 0.51840825, "tRecall_50": 0.55625, "tRecall_75": 0.36875, "tRecall_90": 0.2125, "token_acc": 0.9366985, "epoch": 0.97233725, "global_step/max_steps": "2935/3019", "percentage": "97.22%", "elapsed_time": "11h 1m 12s", "remaining_time": "18m 55s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073981} +{"loss": 0.14085693, "grad_norm": 0.4459908, "learning_rate": 2e-08, "tIoU": 0.57514629, "tRecall_50": 0.625, "tRecall_75": 0.4625, "tRecall_90": 0.33125, "token_acc": 0.94837476, "epoch": 0.97399371, "global_step/max_steps": "2940/3019", "percentage": "97.38%", "elapsed_time": "11h 2m 19s", "remaining_time": "17m 47s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073982} +{"loss": 0.16132812, "grad_norm": 0.62278062, "learning_rate": 2e-08, "tIoU": 0.5526155, "tRecall_50": 0.60625, "tRecall_75": 0.40625, "tRecall_90": 0.28125, "token_acc": 0.94223433, "epoch": 0.97565016, "global_step/max_steps": "2945/3019", "percentage": "97.55%", "elapsed_time": "11h 3m 26s", "remaining_time": "16m 40s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073983} +{"loss": 0.15426025, "grad_norm": 0.54165953, "learning_rate": 1e-08, "tIoU": 0.54847579, "tRecall_50": 0.625, "tRecall_75": 0.43125, "tRecall_90": 0.24375, "token_acc": 0.94067797, "epoch": 0.97730661, "global_step/max_steps": "2950/3019", "percentage": "97.71%", "elapsed_time": "11h 4m 29s", "remaining_time": "15m 32s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073991} +{"loss": 0.16276855, "grad_norm": 1.00697112, "learning_rate": 1e-08, "tIoU": 0.51586361, "tRecall_50": 0.575, "tRecall_75": 0.35625, "tRecall_90": 0.25625, "token_acc": 0.93382554, "epoch": 0.97896306, "global_step/max_steps": "2955/3019", "percentage": "97.88%", "elapsed_time": "11h 5m 34s", "remaining_time": "14m 24s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073996} +{"loss": 0.1557251, "grad_norm": 0.7520166, "learning_rate": 1e-08, "tIoU": 0.55398765, "tRecall_50": 0.58125, "tRecall_75": 0.4, "tRecall_90": 0.29375, "token_acc": 0.94553914, "epoch": 0.98061951, "global_step/max_steps": "2960/3019", "percentage": "98.05%", "elapsed_time": "11h 6m 45s", "remaining_time": "13m 17s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.07399} +{"loss": 0.14675293, "grad_norm": 0.47164127, "learning_rate": 1e-08, "tIoU": 0.57729712, "tRecall_50": 0.625, "tRecall_75": 0.4125, "tRecall_90": 0.30625, "token_acc": 0.94465649, "epoch": 0.98227596, "global_step/max_steps": "2965/3019", "percentage": "98.21%", "elapsed_time": "11h 7m 54s", "remaining_time": "12m 9s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073987} +{"loss": 0.16009521, "grad_norm": 0.58690757, "learning_rate": 1e-08, "tIoU": 0.54488754, "tRecall_50": 0.60625, "tRecall_75": 0.4, "tRecall_90": 0.28125, "token_acc": 0.93994565, "epoch": 0.98393242, "global_step/max_steps": "2970/3019", "percentage": "98.38%", "elapsed_time": "11h 9m 7s", "remaining_time": "11m 2s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073978} +{"loss": 0.15915527, "grad_norm": 0.71665919, "learning_rate": 1e-08, "tIoU": 0.50798087, "tRecall_50": 0.5375, "tRecall_75": 0.3875, "tRecall_90": 0.21875, "token_acc": 0.93811341, "epoch": 0.98558887, "global_step/max_steps": "2975/3019", "percentage": "98.54%", "elapsed_time": "11h 10m 16s", "remaining_time": "9m 54s", "memory(GiB)": 39.98, "train_speed(iter/s)": 0.073975} +{"loss": 0.14058838, "grad_norm": 0.47537214, "learning_rate": 0.0, "tIoU": 0.55793653, "tRecall_50": 0.6, "tRecall_75": 0.43125, "tRecall_90": 0.325, "token_acc": 0.94619565, "epoch": 0.98724532, "global_step/max_steps": "2980/3019", "percentage": "98.71%", "elapsed_time": "11h 11m 24s", "remaining_time": "8m 47s", "memory(GiB)": 40.1, "train_speed(iter/s)": 0.073973} +{"loss": 0.16306152, "grad_norm": 0.63539022, "learning_rate": 0.0, "tIoU": 0.56268797, "tRecall_50": 0.60625, "tRecall_75": 0.3875, "tRecall_90": 0.24375, "token_acc": 0.9439913, "epoch": 0.98890177, "global_step/max_steps": "2985/3019", "percentage": "98.87%", "elapsed_time": "11h 12m 29s", "remaining_time": "7m 39s", "memory(GiB)": 40.1, "train_speed(iter/s)": 0.073978} +{"loss": 0.16289063, "grad_norm": 0.73564398, "learning_rate": 0.0, "tIoU": 0.50330553, "tRecall_50": 0.5625, "tRecall_75": 0.36875, "tRecall_90": 0.20625, "token_acc": 0.93920393, "epoch": 0.99055822, "global_step/max_steps": "2990/3019", "percentage": "99.04%", "elapsed_time": "11h 13m 37s", "remaining_time": "6m 32s", "memory(GiB)": 40.1, "train_speed(iter/s)": 0.073978} +{"loss": 0.16402588, "grad_norm": 0.4798955, "learning_rate": 0.0, "tIoU": 0.51453605, "tRecall_50": 0.55625, "tRecall_75": 0.33125, "tRecall_90": 0.25, "token_acc": 0.93978202, "epoch": 0.99221468, "global_step/max_steps": "2995/3019", "percentage": "99.21%", "elapsed_time": "11h 14m 45s", "remaining_time": "5m 24s", "memory(GiB)": 40.1, "train_speed(iter/s)": 0.073977} +{"loss": 0.15085449, "grad_norm": 0.87243956, "learning_rate": 0.0, "tIoU": 0.58280926, "tRecall_50": 0.65625, "tRecall_75": 0.4375, "tRecall_90": 0.2625, "token_acc": 0.94776931, "epoch": 0.99387113, "global_step/max_steps": "3000/3019", "percentage": "99.37%", "elapsed_time": "11h 15m 56s", "remaining_time": "4m 16s", "memory(GiB)": 40.1, "train_speed(iter/s)": 0.07397} +{"loss": 0.16117554, "grad_norm": 0.44697672, "learning_rate": 0.0, "tIoU": 0.51440773, "tRecall_50": 0.5, "tRecall_75": 0.375, "tRecall_90": 0.275, "token_acc": 0.9394844, "epoch": 0.99552758, "global_step/max_steps": "3005/3019", "percentage": "99.54%", "elapsed_time": "11h 17m 2s", "remaining_time": "3m 9s", "memory(GiB)": 40.1, "train_speed(iter/s)": 0.073974} +{"loss": 0.17885132, "grad_norm": 0.65391296, "learning_rate": 0.0, "tIoU": 0.49717269, "tRecall_50": 0.5375, "tRecall_75": 0.35, "tRecall_90": 0.19375, "token_acc": 0.93235534, "epoch": 0.99718403, "global_step/max_steps": "3010/3019", "percentage": "99.70%", "elapsed_time": "11h 18m 1s", "remaining_time": "2m 1s", "memory(GiB)": 40.1, "train_speed(iter/s)": 0.073988} +{"loss": 0.17543945, "grad_norm": 0.34182894, "learning_rate": 0.0, "tIoU": 0.51140485, "tRecall_50": 0.54375, "tRecall_75": 0.36875, "tRecall_90": 0.26875, "token_acc": 0.93530851, "epoch": 0.99884048, "global_step/max_steps": "3015/3019", "percentage": "99.87%", "elapsed_time": "11h 18m 58s", "remaining_time": "54s", "memory(GiB)": 40.1, "train_speed(iter/s)": 0.074008} +{"train_runtime": 40919.318, "train_samples_per_second": 2.36, "train_steps_per_second": 0.074, "total_flos": 8.424853533524125e+19, "train_loss": 0.18924242, "tIoU": 0.56815686, "tRecall_50": 0.61607143, "tRecall_75": 0.42857143, "tRecall_90": 0.25, "token_acc": 0.94338843, "epoch": 1.0, "global_step/max_steps": "3019/3019", "percentage": "100.00%", "elapsed_time": "11h 21m 57s", "remaining_time": "0s", "memory(GiB)": 40.1, "train_speed(iter/s)": 0.073783} +{"model_parameter_info": "Qwen3VLForConditionalGeneration: 4437.8158M Params (4131.5732M Trainable [93.0993%]), 0.0001M Buffers.", "last_model_checkpoint": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/qwen3vl_4b_sft_timelens_100k_1epoch/v3-20251218-165442/checkpoint-3019", "best_model_checkpoint": null, "best_metric": null, "global_step": 3019, "log_history": [{"loss": 3.310546875, "grad_norm": 37.5207633972168, "learning_rate": 1.098901098901099e-07, "tIoU": 0.0, "tRecall_50": 0.0, "tRecall_75": 0.0, "tRecall_90": 0.0, "token_acc": 0.7585733882030178, "epoch": 0.00033129037601457677, "step": 1}, {"loss": 3.28759765625, "grad_norm": 33.392704010009766, "learning_rate": 5.494505494505495e-07, "tIoU": 0.0, "tRecall_50": 0.0, "tRecall_75": 0.0, "tRecall_90": 0.0, "token_acc": 0.7540308747855917, "epoch": 0.001656451880072884, "step": 5}, {"loss": 2.9546875, "grad_norm": 23.32330894470215, "learning_rate": 1.098901098901099e-06, "tIoU": 0.0, "tRecall_50": 0.0, "tRecall_75": 0.0, "tRecall_90": 0.0, "token_acc": 0.779031370390754, "epoch": 0.003312903760145768, "step": 10}, {"loss": 1.830859375, "grad_norm": 15.744415283203125, "learning_rate": 1.6483516483516484e-06, "tIoU": 0.0, "tRecall_50": 0.0, "tRecall_75": 0.0, "tRecall_90": 0.0, "token_acc": 0.8237547892720306, "epoch": 0.0049693556402186515, "step": 15}, {"loss": 0.6703125, "grad_norm": 13.483047485351562, "learning_rate": 2.197802197802198e-06, "tIoU": 0.15047523975372315, "tRecall_50": 0.15, "tRecall_75": 0.09375, "tRecall_90": 0.05, "token_acc": 0.8428221193135386, "epoch": 0.006625807520291536, "step": 20}, {"loss": 0.21407470703125, "grad_norm": 1.4258636236190796, "learning_rate": 2.7472527472527476e-06, "tIoU": 0.47949848175048826, "tRecall_50": 0.5, "tRecall_75": 0.2875, "tRecall_90": 0.18125, "token_acc": 0.9279303780255643, "epoch": 0.008282259400364419, "step": 25}, {"loss": 0.1758056640625, "grad_norm": 1.9763015508651733, "learning_rate": 3.2967032967032968e-06, "tIoU": 0.5248600482940674, "tRecall_50": 0.59375, "tRecall_75": 0.35625, "tRecall_90": 0.2125, "token_acc": 0.9358974358974359, "epoch": 0.009938711280437303, "step": 30}, {"loss": 0.19344482421875, "grad_norm": 2.235086679458618, "learning_rate": 3.846153846153847e-06, "tIoU": 0.4807941436767578, "tRecall_50": 0.5125, "tRecall_75": 0.33125, "tRecall_90": 0.225, "token_acc": 0.9296811120196239, "epoch": 0.011595163160510187, "step": 35}, {"loss": 0.1904052734375, "grad_norm": 1.306754231452942, "learning_rate": 4.395604395604396e-06, "tIoU": 0.45671963691711426, "tRecall_50": 0.475, "tRecall_75": 0.275, "tRecall_90": 0.15625, "token_acc": 0.9292107404393816, "epoch": 0.013251615040583071, "step": 40}, {"loss": 0.18565673828125, "grad_norm": 1.1737031936645508, "learning_rate": 4.945054945054946e-06, "tIoU": 0.4578300952911377, "tRecall_50": 0.5, "tRecall_75": 0.34375, "tRecall_90": 0.24375, "token_acc": 0.9360131255127153, "epoch": 0.014908066920655955, "step": 45}, {"loss": 0.19178466796875, "grad_norm": 1.6940656900405884, "learning_rate": 5.494505494505495e-06, "tIoU": 0.44655814170837405, "tRecall_50": 0.44375, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.9292543021032504, "epoch": 0.016564518800728838, "step": 50}, {"loss": 0.18096923828125, "grad_norm": 2.1192846298217773, "learning_rate": 6.043956043956044e-06, "tIoU": 0.44701719284057617, "tRecall_50": 0.5, "tRecall_75": 0.31875, "tRecall_90": 0.16875, "token_acc": 0.9279824079164376, "epoch": 0.018220970680801724, "step": 55}, {"loss": 0.210260009765625, "grad_norm": 1.134567141532898, "learning_rate": 6.5934065934065935e-06, "tIoU": 0.39896743297576903, "tRecall_50": 0.4125, "tRecall_75": 0.30625, "tRecall_90": 0.19375, "token_acc": 0.925845147219193, "epoch": 0.019877422560874606, "step": 60}, {"loss": 0.19576416015625, "grad_norm": 2.079129219055176, "learning_rate": 7.1428571428571436e-06, "tIoU": 0.4517733097076416, "tRecall_50": 0.51875, "tRecall_75": 0.26875, "tRecall_90": 0.14375, "token_acc": 0.9273318872017353, "epoch": 0.021533874440947492, "step": 65}, {"loss": 0.198193359375, "grad_norm": 2.3562278747558594, "learning_rate": 7.692307692307694e-06, "tIoU": 0.48453369140625, "tRecall_50": 0.53125, "tRecall_75": 0.35625, "tRecall_90": 0.19375, "token_acc": 0.9291166848418757, "epoch": 0.023190326321020374, "step": 70}, {"loss": 0.1886962890625, "grad_norm": 0.9555505514144897, "learning_rate": 8.241758241758243e-06, "tIoU": 0.4445176601409912, "tRecall_50": 0.4875, "tRecall_75": 0.3375, "tRecall_90": 0.1625, "token_acc": 0.9272082878953108, "epoch": 0.024846778201093257, "step": 75}, {"loss": 0.18936767578125, "grad_norm": 1.0119189023971558, "learning_rate": 8.791208791208792e-06, "tIoU": 0.4313098907470703, "tRecall_50": 0.46875, "tRecall_75": 0.25625, "tRecall_90": 0.15, "token_acc": 0.9278518921862238, "epoch": 0.026503230081166142, "step": 80}, {"loss": 0.2029052734375, "grad_norm": 1.2861628532409668, "learning_rate": 9.340659340659341e-06, "tIoU": 0.42988147735595705, "tRecall_50": 0.44375, "tRecall_75": 0.3125, "tRecall_90": 0.19375, "token_acc": 0.9288451012588944, "epoch": 0.028159681961239025, "step": 85}, {"loss": 0.18587646484375, "grad_norm": 0.9592366218566895, "learning_rate": 9.890109890109892e-06, "tIoU": 0.4889820098876953, "tRecall_50": 0.525, "tRecall_75": 0.36875, "tRecall_90": 0.1875, "token_acc": 0.9345156889495225, "epoch": 0.02981613384131191, "step": 90}, {"loss": 0.18406982421875, "grad_norm": 1.294869303703308, "learning_rate": 9.99995395134759e-06, "tIoU": 0.44582033157348633, "tRecall_50": 0.49375, "tRecall_75": 0.30625, "tRecall_90": 0.14375, "token_acc": 0.9277762693456422, "epoch": 0.031472585721384796, "step": 95}, {"loss": 0.209716796875, "grad_norm": 1.088860034942627, "learning_rate": 9.999766880150859e-06, "tIoU": 0.4009531497955322, "tRecall_50": 0.425, "tRecall_75": 0.21875, "tRecall_90": 0.1375, "token_acc": 0.9226368836829202, "epoch": 0.033129037601457675, "step": 100}, {"loss": 0.210595703125, "grad_norm": 0.7980255484580994, "learning_rate": 9.99943591374887e-06, "tIoU": 0.42629294395446776, "tRecall_50": 0.4625, "tRecall_75": 0.275, "tRecall_90": 0.10625, "token_acc": 0.9228687415426252, "epoch": 0.03478548948153056, "step": 105}, {"loss": 0.20194091796875, "grad_norm": 1.2525278329849243, "learning_rate": 9.998961061666965e-06, "tIoU": 0.4068713665008545, "tRecall_50": 0.4375, "tRecall_75": 0.2375, "tRecall_90": 0.14375, "token_acc": 0.9252514270182115, "epoch": 0.03644194136160345, "step": 110}, {"loss": 0.2082275390625, "grad_norm": 1.621291995048523, "learning_rate": 9.998342337571566e-06, "tIoU": 0.4473285675048828, "tRecall_50": 0.50625, "tRecall_75": 0.2625, "tRecall_90": 0.14375, "token_acc": 0.9251904243743199, "epoch": 0.038098393241676326, "step": 115}, {"loss": 0.2082763671875, "grad_norm": 0.8591983914375305, "learning_rate": 9.997579759269792e-06, "tIoU": 0.419110631942749, "tRecall_50": 0.44375, "tRecall_75": 0.29375, "tRecall_90": 0.15625, "token_acc": 0.92253136933988, "epoch": 0.03975484512174921, "step": 120}, {"loss": 0.203955078125, "grad_norm": 4.022842884063721, "learning_rate": 9.996673348708936e-06, "tIoU": 0.41852197647094724, "tRecall_50": 0.4375, "tRecall_75": 0.26875, "tRecall_90": 0.14375, "token_acc": 0.9234527687296417, "epoch": 0.0414112970018221, "step": 125}, {"loss": 0.2289794921875, "grad_norm": 1.0151914358139038, "learning_rate": 9.995623131975842e-06, "tIoU": 0.3222654342651367, "tRecall_50": 0.35, "tRecall_75": 0.21875, "tRecall_90": 0.125, "token_acc": 0.9190579317812669, "epoch": 0.043067748881894984, "step": 130}, {"loss": 0.2308349609375, "grad_norm": 0.762208104133606, "learning_rate": 9.99442913929615e-06, "tIoU": 0.348409628868103, "tRecall_50": 0.375, "tRecall_75": 0.2, "tRecall_90": 0.10625, "token_acc": 0.9160076356694846, "epoch": 0.04472420076196786, "step": 135}, {"loss": 0.2153564453125, "grad_norm": 1.0656577348709106, "learning_rate": 9.993091405033427e-06, "tIoU": 0.34590022563934325, "tRecall_50": 0.34375, "tRecall_75": 0.21875, "tRecall_90": 0.15625, "token_acc": 0.9209518599562363, "epoch": 0.04638065264204075, "step": 140}, {"loss": 0.1908203125, "grad_norm": 1.2896987199783325, "learning_rate": 9.991609967688177e-06, "tIoU": 0.468013858795166, "tRecall_50": 0.5125, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.9322774440196614, "epoch": 0.048037104522113634, "step": 145}, {"loss": 0.2052490234375, "grad_norm": 0.8637948036193848, "learning_rate": 9.989984869896734e-06, "tIoU": 0.43006191253662107, "tRecall_50": 0.43125, "tRecall_75": 0.2875, "tRecall_90": 0.14375, "token_acc": 0.9274017467248908, "epoch": 0.04969355640218651, "step": 150}, {"loss": 0.18021240234375, "grad_norm": 1.6045029163360596, "learning_rate": 9.988216158430033e-06, "tIoU": 0.5069736480712891, "tRecall_50": 0.55625, "tRecall_75": 0.4, "tRecall_90": 0.21875, "token_acc": 0.9343860604410563, "epoch": 0.0513500082822594, "step": 155}, {"loss": 0.19678955078125, "grad_norm": 1.5454106330871582, "learning_rate": 9.986303884192269e-06, "tIoU": 0.44966754913330076, "tRecall_50": 0.5, "tRecall_75": 0.3125, "tRecall_90": 0.16875, "token_acc": 0.926550598476605, "epoch": 0.053006460162332285, "step": 160}, {"loss": 0.224371337890625, "grad_norm": 0.8348360657691956, "learning_rate": 9.984248102219428e-06, "tIoU": 0.4493142604827881, "tRecall_50": 0.475, "tRecall_75": 0.30625, "tRecall_90": 0.1875, "token_acc": 0.9201303993480032, "epoch": 0.05466291204240517, "step": 165}, {"loss": 0.199365234375, "grad_norm": 1.0888898372650146, "learning_rate": 9.9820488716777e-06, "tIoU": 0.434508752822876, "tRecall_50": 0.45625, "tRecall_75": 0.2625, "tRecall_90": 0.13125, "token_acc": 0.9254880395930712, "epoch": 0.05631936392247805, "step": 170}, {"loss": 0.1918701171875, "grad_norm": 1.4940712451934814, "learning_rate": 9.97970625586178e-06, "tIoU": 0.4550649166107178, "tRecall_50": 0.49375, "tRecall_75": 0.3375, "tRecall_90": 0.15625, "token_acc": 0.9296427597491137, "epoch": 0.057975815802550935, "step": 175}, {"loss": 0.2085205078125, "grad_norm": 0.890186071395874, "learning_rate": 9.977220322193056e-06, "tIoU": 0.3760667324066162, "tRecall_50": 0.40625, "tRecall_75": 0.24375, "tRecall_90": 0.11875, "token_acc": 0.9240851993446204, "epoch": 0.05963226768262382, "step": 180}, {"loss": 0.19017333984375, "grad_norm": 0.7507341504096985, "learning_rate": 9.974591142217643e-06, "tIoU": 0.46780953407287595, "tRecall_50": 0.525, "tRecall_75": 0.3375, "tRecall_90": 0.15, "token_acc": 0.9292349726775956, "epoch": 0.06128871956269671, "step": 185}, {"loss": 0.23975830078125, "grad_norm": 1.2498581409454346, "learning_rate": 9.971818791604352e-06, "tIoU": 0.37864179611206056, "tRecall_50": 0.38125, "tRecall_75": 0.24375, "tRecall_90": 0.125, "token_acc": 0.9137884144683166, "epoch": 0.06294517144276959, "step": 190}, {"loss": 0.192022705078125, "grad_norm": 1.1122004985809326, "learning_rate": 9.968903350142494e-06, "tIoU": 0.4604650020599365, "tRecall_50": 0.46875, "tRecall_75": 0.3375, "tRecall_90": 0.2, "token_acc": 0.9331331877729258, "epoch": 0.06460162332284247, "step": 195}, {"loss": 0.2272705078125, "grad_norm": 0.8076160550117493, "learning_rate": 9.965844901739587e-06, "tIoU": 0.32807245254516604, "tRecall_50": 0.35, "tRecall_75": 0.2, "tRecall_90": 0.14375, "token_acc": 0.9176663031624863, "epoch": 0.06625807520291535, "step": 200}, {"loss": 0.2008056640625, "grad_norm": 1.5814825296401978, "learning_rate": 9.962643534418954e-06, "tIoU": 0.4125843524932861, "tRecall_50": 0.475, "tRecall_75": 0.2375, "tRecall_90": 0.15, "token_acc": 0.9271089271089271, "epoch": 0.06791452708298824, "step": 205}, {"loss": 0.19127197265625, "grad_norm": 0.9881187677383423, "learning_rate": 9.959299340317167e-06, "tIoU": 0.4517991065979004, "tRecall_50": 0.5, "tRecall_75": 0.28125, "tRecall_90": 0.13125, "token_acc": 0.9266104919815167, "epoch": 0.06957097896306112, "step": 210}, {"loss": 0.20115966796875, "grad_norm": 0.862718403339386, "learning_rate": 9.955812415681415e-06, "tIoU": 0.43936910629272463, "tRecall_50": 0.44375, "tRecall_75": 0.26875, "tRecall_90": 0.15625, "token_acc": 0.925531914893617, "epoch": 0.071227430843134, "step": 215}, {"loss": 0.206103515625, "grad_norm": 0.8651227951049805, "learning_rate": 9.952182860866723e-06, "tIoU": 0.4165684700012207, "tRecall_50": 0.4125, "tRecall_75": 0.23125, "tRecall_90": 0.1375, "token_acc": 0.9257033597377765, "epoch": 0.0728838827232069, "step": 220}, {"loss": 0.20802001953125, "grad_norm": 0.7443090081214905, "learning_rate": 9.94841078033307e-06, "tIoU": 0.41522841453552245, "tRecall_50": 0.425, "tRecall_75": 0.30625, "tRecall_90": 0.1625, "token_acc": 0.924877517691889, "epoch": 0.07454033460327977, "step": 225}, {"loss": 0.19639892578125, "grad_norm": 0.6482592821121216, "learning_rate": 9.944496282642377e-06, "tIoU": 0.4244387626647949, "tRecall_50": 0.45, "tRecall_75": 0.3, "tRecall_90": 0.1625, "token_acc": 0.9295890410958904, "epoch": 0.07619678648335265, "step": 230}, {"loss": 0.1994873046875, "grad_norm": 0.8621935248374939, "learning_rate": 9.940439480455386e-06, "tIoU": 0.43390598297119143, "tRecall_50": 0.4625, "tRecall_75": 0.28125, "tRecall_90": 0.14375, "token_acc": 0.9281842818428184, "epoch": 0.07785323836342554, "step": 235}, {"loss": 0.2104248046875, "grad_norm": 0.9681967496871948, "learning_rate": 9.936240490528419e-06, "tIoU": 0.3532999277114868, "tRecall_50": 0.35, "tRecall_75": 0.2, "tRecall_90": 0.1125, "token_acc": 0.9234345091605141, "epoch": 0.07950969024349842, "step": 240}, {"loss": 0.2027099609375, "grad_norm": 2.0264437198638916, "learning_rate": 9.931899433710011e-06, "tIoU": 0.40817694664001464, "tRecall_50": 0.43125, "tRecall_75": 0.275, "tRecall_90": 0.2, "token_acc": 0.924863387978142, "epoch": 0.08116614212357132, "step": 245}, {"loss": 0.2103515625, "grad_norm": 1.126722812652588, "learning_rate": 9.927416434937442e-06, "tIoU": 0.4178160190582275, "tRecall_50": 0.4625, "tRecall_75": 0.2875, "tRecall_90": 0.1625, "token_acc": 0.9214285714285714, "epoch": 0.0828225940036442, "step": 250}, {"loss": 0.20733642578125, "grad_norm": 0.7986385226249695, "learning_rate": 9.922791623233134e-06, "tIoU": 0.45475945472717283, "tRecall_50": 0.49375, "tRecall_75": 0.30625, "tRecall_90": 0.20625, "token_acc": 0.9269493844049248, "epoch": 0.08447904588371707, "step": 255}, {"loss": 0.2189697265625, "grad_norm": 0.7802565693855286, "learning_rate": 9.918025131700935e-06, "tIoU": 0.390622615814209, "tRecall_50": 0.43125, "tRecall_75": 0.225, "tRecall_90": 0.125, "token_acc": 0.9224511930585684, "epoch": 0.08613549776378997, "step": 260}, {"loss": 0.21519775390625, "grad_norm": 0.6944566369056702, "learning_rate": 9.9131170975223e-06, "tIoU": 0.37258186340332033, "tRecall_50": 0.40625, "tRecall_75": 0.23125, "tRecall_90": 0.1125, "token_acc": 0.9179165530406327, "epoch": 0.08779194964386285, "step": 265}, {"loss": 0.23343505859375, "grad_norm": 0.7498219609260559, "learning_rate": 9.90806766195233e-06, "tIoU": 0.3832035779953003, "tRecall_50": 0.39375, "tRecall_75": 0.24375, "tRecall_90": 0.1375, "token_acc": 0.913684785577711, "epoch": 0.08944840152393572, "step": 270}, {"loss": 0.20540771484375, "grad_norm": 0.996425211429596, "learning_rate": 9.902876970315715e-06, "tIoU": 0.44122681617736814, "tRecall_50": 0.5, "tRecall_75": 0.25, "tRecall_90": 0.1375, "token_acc": 0.9231607629427793, "epoch": 0.09110485340400862, "step": 275}, {"loss": 0.2068359375, "grad_norm": 0.9214315414428711, "learning_rate": 9.897545172002547e-06, "tIoU": 0.42021684646606444, "tRecall_50": 0.4375, "tRecall_75": 0.29375, "tRecall_90": 0.1625, "token_acc": 0.9262295081967213, "epoch": 0.0927613052840815, "step": 280}, {"loss": 0.197119140625, "grad_norm": 3.090954065322876, "learning_rate": 9.892072420464026e-06, "tIoU": 0.43810601234436036, "tRecall_50": 0.48125, "tRecall_75": 0.2875, "tRecall_90": 0.16875, "token_acc": 0.9290605313612709, "epoch": 0.09441775716415438, "step": 285}, {"loss": 0.20242919921875, "grad_norm": 1.0458402633666992, "learning_rate": 9.886458873208032e-06, "tIoU": 0.44785561561584475, "tRecall_50": 0.5, "tRecall_75": 0.30625, "tRecall_90": 0.1875, "token_acc": 0.9299154622307063, "epoch": 0.09607420904422727, "step": 290}, {"loss": 0.2062255859375, "grad_norm": 0.8456872701644897, "learning_rate": 9.880704691794608e-06, "tIoU": 0.42615447044372556, "tRecall_50": 0.45, "tRecall_75": 0.3, "tRecall_90": 0.16875, "token_acc": 0.9247135842880524, "epoch": 0.09773066092430015, "step": 295}, {"loss": 0.20479736328125, "grad_norm": 0.9653769135475159, "learning_rate": 9.874810041831298e-06, "tIoU": 0.39232654571533204, "tRecall_50": 0.4125, "tRecall_75": 0.28125, "tRecall_90": 0.14375, "token_acc": 0.9242053789731052, "epoch": 0.09938711280437303, "step": 300}, {"loss": 0.1989013671875, "grad_norm": 0.8343780636787415, "learning_rate": 9.868775092968384e-06, "tIoU": 0.41749234199523927, "tRecall_50": 0.4375, "tRecall_75": 0.3125, "tRecall_90": 0.15, "token_acc": 0.9305402425578831, "epoch": 0.10104356468444592, "step": 305}, {"loss": 0.19197998046875, "grad_norm": 0.815432608127594, "learning_rate": 9.862600018894004e-06, "tIoU": 0.4726102828979492, "tRecall_50": 0.55, "tRecall_75": 0.3125, "tRecall_90": 0.15, "token_acc": 0.9284158954817637, "epoch": 0.1027000165645188, "step": 310}, {"loss": 0.20172119140625, "grad_norm": 0.6193230152130127, "learning_rate": 9.856284997329158e-06, "tIoU": 0.4494438171386719, "tRecall_50": 0.49375, "tRecall_75": 0.3375, "tRecall_90": 0.1875, "token_acc": 0.9278350515463918, "epoch": 0.10435646844459169, "step": 315}, {"loss": 0.20203857421875, "grad_norm": 0.7088842391967773, "learning_rate": 9.849830210022582e-06, "tIoU": 0.43042378425598143, "tRecall_50": 0.45625, "tRecall_75": 0.29375, "tRecall_90": 0.15625, "token_acc": 0.9257748776508973, "epoch": 0.10601292032466457, "step": 320}, {"loss": 0.196875, "grad_norm": 0.6478131413459778, "learning_rate": 9.843235842745527e-06, "tIoU": 0.45046267509460447, "tRecall_50": 0.5, "tRecall_75": 0.3375, "tRecall_90": 0.15625, "token_acc": 0.9284739284739285, "epoch": 0.10766937220473745, "step": 325}, {"loss": 0.20408935546875, "grad_norm": 0.5544605851173401, "learning_rate": 9.836502085286408e-06, "tIoU": 0.47177977561950685, "tRecall_50": 0.5, "tRecall_75": 0.33125, "tRecall_90": 0.19375, "token_acc": 0.9275599128540305, "epoch": 0.10932582408481034, "step": 330}, {"loss": 0.202001953125, "grad_norm": 0.7553025484085083, "learning_rate": 9.829629131445342e-06, "tIoU": 0.424683141708374, "tRecall_50": 0.45, "tRecall_75": 0.25625, "tRecall_90": 0.1375, "token_acc": 0.9231816943612094, "epoch": 0.11098227596488322, "step": 335}, {"loss": 0.3250732421875, "grad_norm": 24.649581909179688, "learning_rate": 9.822617179028575e-06, "tIoU": 0.33823087215423586, "tRecall_50": 0.3625, "tRecall_75": 0.23125, "tRecall_90": 0.16875, "token_acc": 0.9176630434782609, "epoch": 0.1126387278449561, "step": 340}, {"loss": 0.19600830078125, "grad_norm": 1.2221829891204834, "learning_rate": 9.815466429842781e-06, "tIoU": 0.4647655487060547, "tRecall_50": 0.51875, "tRecall_75": 0.34375, "tRecall_90": 0.18125, "token_acc": 0.9318860244233379, "epoch": 0.11429517972502899, "step": 345}, {"loss": 0.19267578125, "grad_norm": 0.5191197395324707, "learning_rate": 9.808177089689258e-06, "tIoU": 0.444231653213501, "tRecall_50": 0.49375, "tRecall_75": 0.28125, "tRecall_90": 0.1375, "token_acc": 0.9275834011391375, "epoch": 0.11595163160510187, "step": 350}, {"loss": 0.19835205078125, "grad_norm": 0.7791886925697327, "learning_rate": 9.80074936835801e-06, "tIoU": 0.43950557708740234, "tRecall_50": 0.425, "tRecall_75": 0.2875, "tRecall_90": 0.15, "token_acc": 0.9243789243789243, "epoch": 0.11760808348517475, "step": 355}, {"loss": 0.2083251953125, "grad_norm": 0.6446515917778015, "learning_rate": 9.793183479621698e-06, "tIoU": 0.4317431926727295, "tRecall_50": 0.45625, "tRecall_75": 0.25625, "tRecall_90": 0.15625, "token_acc": 0.9247691472026073, "epoch": 0.11926453536524764, "step": 360}, {"loss": 0.19989013671875, "grad_norm": 0.6463059186935425, "learning_rate": 9.7854796412295e-06, "tIoU": 0.4000588893890381, "tRecall_50": 0.43125, "tRecall_75": 0.21875, "tRecall_90": 0.125, "token_acc": 0.9270691068014204, "epoch": 0.12092098724532052, "step": 365}, {"loss": 0.20377197265625, "grad_norm": 0.8697914481163025, "learning_rate": 9.777638074900831e-06, "tIoU": 0.38347530364990234, "tRecall_50": 0.4, "tRecall_75": 0.26875, "tRecall_90": 0.1375, "token_acc": 0.9237913138486752, "epoch": 0.12257743912539341, "step": 370}, {"loss": 0.19049072265625, "grad_norm": 0.8284282684326172, "learning_rate": 9.769659006318976e-06, "tIoU": 0.41577610969543455, "tRecall_50": 0.43125, "tRecall_75": 0.30625, "tRecall_90": 0.2125, "token_acc": 0.933115823817292, "epoch": 0.1242338910054663, "step": 375}, {"loss": 0.19884033203125, "grad_norm": 0.544400155544281, "learning_rate": 9.761542665124581e-06, "tIoU": 0.4304636001586914, "tRecall_50": 0.48125, "tRecall_75": 0.275, "tRecall_90": 0.1625, "token_acc": 0.9262697979246314, "epoch": 0.12589034288553919, "step": 380}, {"loss": 0.196875, "grad_norm": 0.7749924659729004, "learning_rate": 9.753289284909058e-06, "tIoU": 0.45472006797790526, "tRecall_50": 0.4625, "tRecall_75": 0.3, "tRecall_90": 0.18125, "token_acc": 0.9281437125748503, "epoch": 0.12754679476561206, "step": 385}, {"loss": 0.189599609375, "grad_norm": 0.8247610330581665, "learning_rate": 9.744899103207844e-06, "tIoU": 0.46105518341064455, "tRecall_50": 0.45, "tRecall_75": 0.33125, "tRecall_90": 0.16875, "token_acc": 0.9312328767123288, "epoch": 0.12920324664568494, "step": 390}, {"loss": 0.2018798828125, "grad_norm": 1.2671105861663818, "learning_rate": 9.736372361493584e-06, "tIoU": 0.41598782539367674, "tRecall_50": 0.4625, "tRecall_75": 0.2625, "tRecall_90": 0.125, "token_acc": 0.9266104919815167, "epoch": 0.13085969852575782, "step": 395}, {"loss": 0.20133056640625, "grad_norm": 0.7221733331680298, "learning_rate": 9.72770930516917e-06, "tIoU": 0.43442955017089846, "tRecall_50": 0.43125, "tRecall_75": 0.275, "tRecall_90": 0.1625, "token_acc": 0.9219219219219219, "epoch": 0.1325161504058307, "step": 400}, {"loss": 0.19154052734375, "grad_norm": 0.8075146079063416, "learning_rate": 9.718910183560681e-06, "tIoU": 0.4739804744720459, "tRecall_50": 0.53125, "tRecall_75": 0.31875, "tRecall_90": 0.14375, "token_acc": 0.9313725490196079, "epoch": 0.1341726022859036, "step": 405}, {"loss": 0.18717041015625, "grad_norm": 0.6628204584121704, "learning_rate": 9.709975249910203e-06, "tIoU": 0.48113312721252444, "tRecall_50": 0.51875, "tRecall_75": 0.33125, "tRecall_90": 0.20625, "token_acc": 0.9319486198414867, "epoch": 0.1358290541659765, "step": 410}, {"loss": 0.2063232421875, "grad_norm": 0.7093484401702881, "learning_rate": 9.70090476136855e-06, "tIoU": 0.41611781120300295, "tRecall_50": 0.4625, "tRecall_75": 0.275, "tRecall_90": 0.125, "token_acc": 0.9220920730046309, "epoch": 0.13748550604604937, "step": 415}, {"loss": 0.18984375, "grad_norm": 0.5163065791130066, "learning_rate": 9.691698978987855e-06, "tIoU": 0.45911922454833987, "tRecall_50": 0.4875, "tRecall_75": 0.30625, "tRecall_90": 0.19375, "token_acc": 0.931214031241436, "epoch": 0.13914195792612225, "step": 420}, {"loss": 0.1943115234375, "grad_norm": 0.6479618549346924, "learning_rate": 9.68235816771406e-06, "tIoU": 0.41985330581665037, "tRecall_50": 0.45625, "tRecall_75": 0.2375, "tRecall_90": 0.1125, "token_acc": 0.9246873300706906, "epoch": 0.14079840980619512, "step": 425}, {"loss": 0.19625244140625, "grad_norm": 1.3915655612945557, "learning_rate": 9.672882596379289e-06, "tIoU": 0.4425868034362793, "tRecall_50": 0.46875, "tRecall_75": 0.24375, "tRecall_90": 0.125, "token_acc": 0.9271739130434783, "epoch": 0.142454861686268, "step": 430}, {"loss": 0.196826171875, "grad_norm": 0.874627947807312, "learning_rate": 9.663272537694112e-06, "tIoU": 0.4497288703918457, "tRecall_50": 0.45, "tRecall_75": 0.31875, "tRecall_90": 0.1625, "token_acc": 0.9272677744483792, "epoch": 0.1441113135663409, "step": 435}, {"loss": 0.21339111328125, "grad_norm": 0.6494178771972656, "learning_rate": 9.653528268239703e-06, "tIoU": 0.3722111940383911, "tRecall_50": 0.36875, "tRecall_75": 0.225, "tRecall_90": 0.15, "token_acc": 0.9220708446866485, "epoch": 0.1457677654464138, "step": 440}, {"loss": 0.18795166015625, "grad_norm": 0.6357215642929077, "learning_rate": 9.643650068459863e-06, "tIoU": 0.4609790325164795, "tRecall_50": 0.50625, "tRecall_75": 0.30625, "tRecall_90": 0.16875, "token_acc": 0.9316888045540797, "epoch": 0.14742421732648667, "step": 445}, {"loss": 0.19854736328125, "grad_norm": 0.9134014844894409, "learning_rate": 9.633638222652967e-06, "tIoU": 0.4400688648223877, "tRecall_50": 0.5, "tRecall_75": 0.28125, "tRecall_90": 0.15, "token_acc": 0.9271884374147805, "epoch": 0.14908066920655955, "step": 450}, {"loss": 0.18902587890625, "grad_norm": 1.318121314048767, "learning_rate": 9.62349301896377e-06, "tIoU": 0.4806161880493164, "tRecall_50": 0.55, "tRecall_75": 0.31875, "tRecall_90": 0.1875, "token_acc": 0.9309967141292442, "epoch": 0.15073712108663243, "step": 455}, {"loss": 0.200537109375, "grad_norm": 0.8935189247131348, "learning_rate": 9.613214749375117e-06, "tIoU": 0.43247547149658205, "tRecall_50": 0.45625, "tRecall_75": 0.24375, "tRecall_90": 0.1375, "token_acc": 0.9250476969201418, "epoch": 0.1523935729667053, "step": 460}, {"loss": 0.21024169921875, "grad_norm": 0.8290932178497314, "learning_rate": 9.602803709699544e-06, "tIoU": 0.4437929630279541, "tRecall_50": 0.4625, "tRecall_75": 0.275, "tRecall_90": 0.1875, "token_acc": 0.9262295081967213, "epoch": 0.1540500248467782, "step": 465}, {"loss": 0.19241943359375, "grad_norm": 0.7094597816467285, "learning_rate": 9.592260199570761e-06, "tIoU": 0.4886190891265869, "tRecall_50": 0.53125, "tRecall_75": 0.30625, "tRecall_90": 0.19375, "token_acc": 0.9312414733969986, "epoch": 0.1557064767268511, "step": 470}, {"loss": 0.19324951171875, "grad_norm": 0.5024470090866089, "learning_rate": 9.581584522435025e-06, "tIoU": 0.4388664722442627, "tRecall_50": 0.4875, "tRecall_75": 0.30625, "tRecall_90": 0.1875, "token_acc": 0.9317563893420338, "epoch": 0.15736292860692397, "step": 475}, {"loss": 0.2156982421875, "grad_norm": 0.7557206153869629, "learning_rate": 9.570776985542412e-06, "tIoU": 0.4001711368560791, "tRecall_50": 0.43125, "tRecall_75": 0.25625, "tRecall_90": 0.1375, "token_acc": 0.9247751430907605, "epoch": 0.15901938048699685, "step": 480}, {"loss": 0.189794921875, "grad_norm": 0.8227375149726868, "learning_rate": 9.559837899937969e-06, "tIoU": 0.47867345809936523, "tRecall_50": 0.50625, "tRecall_75": 0.3375, "tRecall_90": 0.16875, "token_acc": 0.9310250817884406, "epoch": 0.16067583236706973, "step": 485}, {"loss": 0.19400634765625, "grad_norm": 0.920362114906311, "learning_rate": 9.548767580452774e-06, "tIoU": 0.47590069770812987, "tRecall_50": 0.5375, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.9307441607821836, "epoch": 0.16233228424714263, "step": 490}, {"loss": 0.1809326171875, "grad_norm": 0.8449790477752686, "learning_rate": 9.537566345694858e-06, "tIoU": 0.478501033782959, "tRecall_50": 0.5, "tRecall_75": 0.34375, "tRecall_90": 0.20625, "token_acc": 0.9327342047930284, "epoch": 0.1639887361272155, "step": 495}, {"loss": 0.21148681640625, "grad_norm": 0.7748955488204956, "learning_rate": 9.526234518040045e-06, "tIoU": 0.43302106857299805, "tRecall_50": 0.45, "tRecall_75": 0.28125, "tRecall_90": 0.175, "token_acc": 0.9221980413492927, "epoch": 0.1656451880072884, "step": 500}, {"loss": 0.19722900390625, "grad_norm": 0.6476784944534302, "learning_rate": 9.514772423622675e-06, "tIoU": 0.4332547664642334, "tRecall_50": 0.46875, "tRecall_75": 0.28125, "tRecall_90": 0.1375, "token_acc": 0.928298800436205, "epoch": 0.16730163988736127, "step": 505}, {"loss": 0.19720458984375, "grad_norm": 0.7341322898864746, "learning_rate": 9.503180392326216e-06, "tIoU": 0.47381792068481443, "tRecall_50": 0.4875, "tRecall_75": 0.31875, "tRecall_90": 0.19375, "token_acc": 0.9332971800433839, "epoch": 0.16895809176743415, "step": 510}, {"loss": 0.192437744140625, "grad_norm": 0.9431743621826172, "learning_rate": 9.491458757773766e-06, "tIoU": 0.4060690402984619, "tRecall_50": 0.43125, "tRecall_75": 0.2875, "tRecall_90": 0.15625, "token_acc": 0.92718313714755, "epoch": 0.17061454364750703, "step": 515}, {"loss": 0.19864501953125, "grad_norm": 0.9373649954795837, "learning_rate": 9.479607857318454e-06, "tIoU": 0.4573049068450928, "tRecall_50": 0.4875, "tRecall_75": 0.325, "tRecall_90": 0.16875, "token_acc": 0.9318305268875611, "epoch": 0.17227099552757993, "step": 520}, {"loss": 0.19825439453125, "grad_norm": 0.6910871863365173, "learning_rate": 9.467628032033736e-06, "tIoU": 0.4477407455444336, "tRecall_50": 0.45625, "tRecall_75": 0.33125, "tRecall_90": 0.16875, "token_acc": 0.9314505072662462, "epoch": 0.1739274474076528, "step": 525}, {"loss": 0.199853515625, "grad_norm": 0.495835542678833, "learning_rate": 9.455519626703569e-06, "tIoU": 0.4348471164703369, "tRecall_50": 0.45625, "tRecall_75": 0.3, "tRecall_90": 0.19375, "token_acc": 0.9249796140255504, "epoch": 0.1755838992877257, "step": 530}, {"loss": 0.19815673828125, "grad_norm": 0.5815764665603638, "learning_rate": 9.443282989812494e-06, "tIoU": 0.41817460060119627, "tRecall_50": 0.44375, "tRecall_75": 0.2625, "tRecall_90": 0.15, "token_acc": 0.9251700680272109, "epoch": 0.17724035116779857, "step": 535}, {"loss": 0.18758544921875, "grad_norm": 0.834153413772583, "learning_rate": 9.430918473535608e-06, "tIoU": 0.41258726119995115, "tRecall_50": 0.45625, "tRecall_75": 0.28125, "tRecall_90": 0.15625, "token_acc": 0.925734494015234, "epoch": 0.17889680304787145, "step": 540}, {"loss": 0.1734130859375, "grad_norm": 0.49157267808914185, "learning_rate": 9.418426433728422e-06, "tIoU": 0.49922776222229004, "tRecall_50": 0.55, "tRecall_75": 0.3875, "tRecall_90": 0.2375, "token_acc": 0.9362452523060227, "epoch": 0.18055325492794436, "step": 545}, {"loss": 0.20540771484375, "grad_norm": 0.8604575991630554, "learning_rate": 9.405807229916627e-06, "tIoU": 0.40863842964172364, "tRecall_50": 0.4375, "tRecall_75": 0.28125, "tRecall_90": 0.1875, "token_acc": 0.9271232876712329, "epoch": 0.18220970680801724, "step": 550}, {"loss": 0.18697509765625, "grad_norm": 0.8063392043113708, "learning_rate": 9.393061225285743e-06, "tIoU": 0.4734005928039551, "tRecall_50": 0.53125, "tRecall_75": 0.3375, "tRecall_90": 0.1875, "token_acc": 0.9273574561403509, "epoch": 0.18386615868809011, "step": 555}, {"loss": 0.1929931640625, "grad_norm": 0.7195932865142822, "learning_rate": 9.380188786670658e-06, "tIoU": 0.47695555686950686, "tRecall_50": 0.49375, "tRecall_75": 0.35, "tRecall_90": 0.23125, "token_acc": 0.9284353741496598, "epoch": 0.185522610568163, "step": 560}, {"loss": 0.2121337890625, "grad_norm": 0.6872114539146423, "learning_rate": 9.367190284545087e-06, "tIoU": 0.3961095571517944, "tRecall_50": 0.4375, "tRecall_75": 0.25625, "tRecall_90": 0.13125, "token_acc": 0.9261285909712722, "epoch": 0.18717906244823587, "step": 565}, {"loss": 0.18876953125, "grad_norm": 0.6529171466827393, "learning_rate": 9.354066093010898e-06, "tIoU": 0.4421760082244873, "tRecall_50": 0.475, "tRecall_75": 0.3125, "tRecall_90": 0.2125, "token_acc": 0.9333150384193194, "epoch": 0.18883551432830875, "step": 570}, {"loss": 0.2117919921875, "grad_norm": 0.9455483555793762, "learning_rate": 9.34081658978735e-06, "tIoU": 0.41131224632263186, "tRecall_50": 0.4375, "tRecall_75": 0.29375, "tRecall_90": 0.1625, "token_acc": 0.9229092890220648, "epoch": 0.19049196620838166, "step": 575}, {"loss": 0.20238037109375, "grad_norm": 0.6294817924499512, "learning_rate": 9.327442156200217e-06, "tIoU": 0.42276864051818847, "tRecall_50": 0.43125, "tRecall_75": 0.2875, "tRecall_90": 0.16875, "token_acc": 0.9247135842880524, "epoch": 0.19214841808845454, "step": 580}, {"loss": 0.19185791015625, "grad_norm": 0.49901363253593445, "learning_rate": 9.31394317717082e-06, "tIoU": 0.4232339859008789, "tRecall_50": 0.46875, "tRecall_75": 0.3125, "tRecall_90": 0.1375, "token_acc": 0.9246463547334058, "epoch": 0.19380486996852742, "step": 585}, {"loss": 0.19619140625, "grad_norm": 0.6557704210281372, "learning_rate": 9.300320041204947e-06, "tIoU": 0.44829621315002444, "tRecall_50": 0.5125, "tRecall_75": 0.275, "tRecall_90": 0.16875, "token_acc": 0.9294503691550451, "epoch": 0.1954613218486003, "step": 590}, {"loss": 0.2063232421875, "grad_norm": 1.5239434242248535, "learning_rate": 9.286573140381663e-06, "tIoU": 0.4620360851287842, "tRecall_50": 0.5125, "tRecall_75": 0.325, "tRecall_90": 0.16875, "token_acc": 0.926655719759168, "epoch": 0.19711777372867317, "step": 595}, {"loss": 0.20166015625, "grad_norm": 0.68989497423172, "learning_rate": 9.27270287034204e-06, "tIoU": 0.44564409255981446, "tRecall_50": 0.49375, "tRecall_75": 0.2875, "tRecall_90": 0.15625, "token_acc": 0.9264705882352942, "epoch": 0.19877422560874605, "step": 600}, {"loss": 0.19608154296875, "grad_norm": 0.8628115057945251, "learning_rate": 9.258709630277758e-06, "tIoU": 0.4197608470916748, "tRecall_50": 0.425, "tRecall_75": 0.29375, "tRecall_90": 0.16875, "token_acc": 0.930574462292404, "epoch": 0.20043067748881896, "step": 605}, {"loss": 0.21024169921875, "grad_norm": 0.5018081068992615, "learning_rate": 9.244593822919627e-06, "tIoU": 0.4352898120880127, "tRecall_50": 0.44375, "tRecall_75": 0.275, "tRecall_90": 0.15625, "token_acc": 0.9244721169463995, "epoch": 0.20208712936889184, "step": 610}, {"loss": 0.1876708984375, "grad_norm": 0.5807973146438599, "learning_rate": 9.230355854525988e-06, "tIoU": 0.45960350036621095, "tRecall_50": 0.475, "tRecall_75": 0.3125, "tRecall_90": 0.19375, "token_acc": 0.931185144729656, "epoch": 0.20374358124896472, "step": 615}, {"loss": 0.2001220703125, "grad_norm": 0.8211379647254944, "learning_rate": 9.215996134871022e-06, "tIoU": 0.478743839263916, "tRecall_50": 0.4875, "tRecall_75": 0.325, "tRecall_90": 0.16875, "token_acc": 0.9288816503800217, "epoch": 0.2054000331290376, "step": 620}, {"loss": 0.19661865234375, "grad_norm": 0.7232769727706909, "learning_rate": 9.201515077232958e-06, "tIoU": 0.4671457290649414, "tRecall_50": 0.5, "tRecall_75": 0.3375, "tRecall_90": 0.16875, "token_acc": 0.9303401360544218, "epoch": 0.20705648500911047, "step": 625}, {"loss": 0.20330810546875, "grad_norm": 0.7774529457092285, "learning_rate": 9.186913098382185e-06, "tIoU": 0.446498441696167, "tRecall_50": 0.46875, "tRecall_75": 0.30625, "tRecall_90": 0.15, "token_acc": 0.924986361156574, "epoch": 0.20871293688918338, "step": 630}, {"loss": 0.18858642578125, "grad_norm": 0.5614256262779236, "learning_rate": 9.172190618569236e-06, "tIoU": 0.4410062313079834, "tRecall_50": 0.48125, "tRecall_75": 0.29375, "tRecall_90": 0.15625, "token_acc": 0.9271232876712329, "epoch": 0.21036938876925626, "step": 635}, {"loss": 0.19208984375, "grad_norm": 0.5532062649726868, "learning_rate": 9.157348061512728e-06, "tIoU": 0.4408290386199951, "tRecall_50": 0.48125, "tRecall_75": 0.2625, "tRecall_90": 0.175, "token_acc": 0.9247751430907605, "epoch": 0.21202584064932914, "step": 640}, {"loss": 0.20494384765625, "grad_norm": 0.6056038737297058, "learning_rate": 9.142385854387128e-06, "tIoU": 0.4316275119781494, "tRecall_50": 0.44375, "tRecall_75": 0.2875, "tRecall_90": 0.18125, "token_acc": 0.9252464403066812, "epoch": 0.21368229252940202, "step": 645}, {"loss": 0.18748779296875, "grad_norm": 0.8482238054275513, "learning_rate": 9.127304427810491e-06, "tIoU": 0.4653031349182129, "tRecall_50": 0.53125, "tRecall_75": 0.3, "tRecall_90": 0.175, "token_acc": 0.9333515432941819, "epoch": 0.2153387444094749, "step": 650}, {"loss": 0.1937255859375, "grad_norm": 0.543383777141571, "learning_rate": 9.112104215832047e-06, "tIoU": 0.4971604824066162, "tRecall_50": 0.5625, "tRecall_75": 0.375, "tRecall_90": 0.19375, "token_acc": 0.9318366242899648, "epoch": 0.21699519628954778, "step": 655}, {"loss": 0.192578125, "grad_norm": 0.6806201338768005, "learning_rate": 9.096785655919717e-06, "tIoU": 0.4546171188354492, "tRecall_50": 0.4875, "tRecall_75": 0.30625, "tRecall_90": 0.1625, "token_acc": 0.9255637055148057, "epoch": 0.21865164816962068, "step": 660}, {"loss": 0.19219970703125, "grad_norm": 0.5829448103904724, "learning_rate": 9.08134918894752e-06, "tIoU": 0.4513428211212158, "tRecall_50": 0.48125, "tRecall_75": 0.29375, "tRecall_90": 0.175, "token_acc": 0.9289395172226743, "epoch": 0.22030810004969356, "step": 665}, {"loss": 0.20467529296875, "grad_norm": 0.8667960166931152, "learning_rate": 9.06579525918289e-06, "tIoU": 0.42314701080322265, "tRecall_50": 0.45, "tRecall_75": 0.2625, "tRecall_90": 0.15625, "token_acc": 0.9248839114995903, "epoch": 0.22196455192976644, "step": 670}, {"loss": 0.192578125, "grad_norm": 0.720363974571228, "learning_rate": 9.050124314273876e-06, "tIoU": 0.46733717918395995, "tRecall_50": 0.4625, "tRecall_75": 0.33125, "tRecall_90": 0.2375, "token_acc": 0.9315143246930423, "epoch": 0.22362100380983932, "step": 675}, {"loss": 0.18389892578125, "grad_norm": 0.6191056966781616, "learning_rate": 9.034336805236279e-06, "tIoU": 0.48230271339416503, "tRecall_50": 0.50625, "tRecall_75": 0.34375, "tRecall_90": 0.23125, "token_acc": 0.9328950972336346, "epoch": 0.2252774556899122, "step": 680}, {"loss": 0.18983154296875, "grad_norm": 1.0231218338012695, "learning_rate": 9.018433186440648e-06, "tIoU": 0.4781657695770264, "tRecall_50": 0.5375, "tRecall_75": 0.31875, "tRecall_90": 0.16875, "token_acc": 0.9283378746594005, "epoch": 0.2269339075699851, "step": 685}, {"loss": 0.19080810546875, "grad_norm": 0.6080626249313354, "learning_rate": 9.002413915599223e-06, "tIoU": 0.42686448097229, "tRecall_50": 0.45, "tRecall_75": 0.2875, "tRecall_90": 0.175, "token_acc": 0.9304109589041096, "epoch": 0.22859035945005798, "step": 690}, {"loss": 0.18778076171875, "grad_norm": 0.6328234076499939, "learning_rate": 8.986279453752758e-06, "tIoU": 0.46008977890014646, "tRecall_50": 0.48125, "tRecall_75": 0.36875, "tRecall_90": 0.20625, "token_acc": 0.9322080043561122, "epoch": 0.23024681133013086, "step": 695}, {"loss": 0.19652099609375, "grad_norm": 0.5854917764663696, "learning_rate": 8.970030265257237e-06, "tIoU": 0.4481083393096924, "tRecall_50": 0.48125, "tRecall_75": 0.3125, "tRecall_90": 0.1875, "token_acc": 0.9253077975376197, "epoch": 0.23190326321020374, "step": 700}, {"loss": 0.184967041015625, "grad_norm": 0.563603401184082, "learning_rate": 8.95366681777053e-06, "tIoU": 0.4827601909637451, "tRecall_50": 0.5, "tRecall_75": 0.3, "tRecall_90": 0.175, "token_acc": 0.9291575492341356, "epoch": 0.23355971509027662, "step": 705}, {"loss": 0.178515625, "grad_norm": 0.6367529034614563, "learning_rate": 8.93718958223892e-06, "tIoU": 0.46496214866638186, "tRecall_50": 0.5375, "tRecall_75": 0.3, "tRecall_90": 0.19375, "token_acc": 0.930536638518115, "epoch": 0.2352161669703495, "step": 710}, {"loss": 0.1771484375, "grad_norm": 0.5359927415847778, "learning_rate": 8.920599032883553e-06, "tIoU": 0.5005170345306397, "tRecall_50": 0.55, "tRecall_75": 0.33125, "tRecall_90": 0.20625, "token_acc": 0.9344620425996724, "epoch": 0.2368726188504224, "step": 715}, {"loss": 0.18304443359375, "grad_norm": 1.3301435708999634, "learning_rate": 8.903895647186795e-06, "tIoU": 0.47890748977661135, "tRecall_50": 0.525, "tRecall_75": 0.3375, "tRecall_90": 0.175, "token_acc": 0.9330966684871654, "epoch": 0.23852907073049529, "step": 720}, {"loss": 0.19580078125, "grad_norm": 0.9849122762680054, "learning_rate": 8.887079905878484e-06, "tIoU": 0.4430722713470459, "tRecall_50": 0.4625, "tRecall_75": 0.29375, "tRecall_90": 0.20625, "token_acc": 0.927891156462585, "epoch": 0.24018552261056816, "step": 725}, {"loss": 0.17957763671875, "grad_norm": 0.4319475293159485, "learning_rate": 8.870152292922088e-06, "tIoU": 0.46705260276794436, "tRecall_50": 0.5125, "tRecall_75": 0.3125, "tRecall_90": 0.19375, "token_acc": 0.9331331877729258, "epoch": 0.24184197449064104, "step": 730}, {"loss": 0.19373779296875, "grad_norm": 0.7241905331611633, "learning_rate": 8.853113295500795e-06, "tIoU": 0.4641860485076904, "tRecall_50": 0.4875, "tRecall_75": 0.29375, "tRecall_90": 0.19375, "token_acc": 0.9301440608861103, "epoch": 0.24349842637071392, "step": 735}, {"loss": 0.18074951171875, "grad_norm": 1.020323395729065, "learning_rate": 8.83596340400347e-06, "tIoU": 0.4771578788757324, "tRecall_50": 0.50625, "tRecall_75": 0.38125, "tRecall_90": 0.2625, "token_acc": 0.934342069710889, "epoch": 0.24515487825078683, "step": 740}, {"loss": 0.19556884765625, "grad_norm": 0.6605811715126038, "learning_rate": 8.818703112010562e-06, "tIoU": 0.4769937992095947, "tRecall_50": 0.51875, "tRecall_75": 0.3375, "tRecall_90": 0.2125, "token_acc": 0.9324104234527687, "epoch": 0.2468113301308597, "step": 745}, {"loss": 0.178759765625, "grad_norm": 0.737177848815918, "learning_rate": 8.801332916279882e-06, "tIoU": 0.4775440216064453, "tRecall_50": 0.4875, "tRecall_75": 0.3375, "tRecall_90": 0.2125, "token_acc": 0.9352028314729104, "epoch": 0.2484677820109326, "step": 750}, {"loss": 0.1890380859375, "grad_norm": 1.0642855167388916, "learning_rate": 8.783853316732313e-06, "tIoU": 0.48411898612976073, "tRecall_50": 0.53125, "tRecall_75": 0.375, "tRecall_90": 0.20625, "token_acc": 0.9313725490196079, "epoch": 0.25012423389100547, "step": 755}, {"loss": 0.17894287109375, "grad_norm": 0.8085300922393799, "learning_rate": 8.766264816437424e-06, "tIoU": 0.4733723163604736, "tRecall_50": 0.4875, "tRecall_75": 0.36875, "tRecall_90": 0.25, "token_acc": 0.9372955288985824, "epoch": 0.25178068577107837, "step": 760}, {"loss": 0.19639892578125, "grad_norm": 1.0604671239852905, "learning_rate": 8.74856792159899e-06, "tIoU": 0.4630789279937744, "tRecall_50": 0.49375, "tRecall_75": 0.325, "tRecall_90": 0.175, "token_acc": 0.9286300246103364, "epoch": 0.2534371376511512, "step": 765}, {"loss": 0.18775634765625, "grad_norm": 0.749838650226593, "learning_rate": 8.730763141540417e-06, "tIoU": 0.46113018989562987, "tRecall_50": 0.5125, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.9296004349007883, "epoch": 0.25509358953122413, "step": 770}, {"loss": 0.1763427734375, "grad_norm": 0.7101616859436035, "learning_rate": 8.712850988690094e-06, "tIoU": 0.4654840469360352, "tRecall_50": 0.50625, "tRecall_75": 0.34375, "tRecall_90": 0.14375, "token_acc": 0.9309309309309309, "epoch": 0.256750041411297, "step": 775}, {"loss": 0.18895263671875, "grad_norm": 0.8688730001449585, "learning_rate": 8.69483197856664e-06, "tIoU": 0.49425182342529295, "tRecall_50": 0.51875, "tRecall_75": 0.3375, "tRecall_90": 0.2, "token_acc": 0.9316146540027137, "epoch": 0.2584064932913699, "step": 780}, {"loss": 0.18460693359375, "grad_norm": 0.4564657211303711, "learning_rate": 8.676706629764066e-06, "tIoU": 0.4841297149658203, "tRecall_50": 0.48125, "tRecall_75": 0.38125, "tRecall_90": 0.24375, "token_acc": 0.9336762317271251, "epoch": 0.2600629451714428, "step": 785}, {"loss": 0.17938232421875, "grad_norm": 0.6564610600471497, "learning_rate": 8.658475463936849e-06, "tIoU": 0.4766863822937012, "tRecall_50": 0.53125, "tRecall_75": 0.3, "tRecall_90": 0.20625, "token_acc": 0.9309594998640935, "epoch": 0.26171939705151565, "step": 790}, {"loss": 0.1773681640625, "grad_norm": 0.587030827999115, "learning_rate": 8.640139005784924e-06, "tIoU": 0.4590290069580078, "tRecall_50": 0.49375, "tRecall_75": 0.31875, "tRecall_90": 0.21875, "token_acc": 0.9338753387533876, "epoch": 0.26337584893158855, "step": 795}, {"loss": 0.19100341796875, "grad_norm": 0.594359815120697, "learning_rate": 8.621697783038579e-06, "tIoU": 0.44837188720703125, "tRecall_50": 0.48125, "tRecall_75": 0.3375, "tRecall_90": 0.175, "token_acc": 0.927969556944822, "epoch": 0.2650323008116614, "step": 800}, {"loss": 0.19122314453125, "grad_norm": 1.1288412809371948, "learning_rate": 8.603152326443262e-06, "tIoU": 0.44898138046264646, "tRecall_50": 0.48125, "tRecall_75": 0.2875, "tRecall_90": 0.2, "token_acc": 0.9313351498637602, "epoch": 0.2666887526917343, "step": 805}, {"loss": 0.18260498046875, "grad_norm": 0.6252028942108154, "learning_rate": 8.584503169744317e-06, "tIoU": 0.4822195529937744, "tRecall_50": 0.5125, "tRecall_75": 0.3125, "tRecall_90": 0.21875, "token_acc": 0.9337513631406761, "epoch": 0.2683452045718072, "step": 810}, {"loss": 0.17353515625, "grad_norm": 0.6308260560035706, "learning_rate": 8.565750849671616e-06, "tIoU": 0.5001877307891845, "tRecall_50": 0.51875, "tRecall_75": 0.38125, "tRecall_90": 0.225, "token_acc": 0.9359149168257431, "epoch": 0.27000165645188007, "step": 815}, {"loss": 0.1913330078125, "grad_norm": 0.9611879587173462, "learning_rate": 8.546895905924108e-06, "tIoU": 0.47109642028808596, "tRecall_50": 0.4875, "tRecall_75": 0.375, "tRecall_90": 0.21875, "token_acc": 0.9290780141843972, "epoch": 0.271658108331953, "step": 820}, {"loss": 0.1817138671875, "grad_norm": 0.6250311732292175, "learning_rate": 8.527938881154298e-06, "tIoU": 0.4798725128173828, "tRecall_50": 0.53125, "tRecall_75": 0.36875, "tRecall_90": 0.1875, "token_acc": 0.9346816069964471, "epoch": 0.2733145602120258, "step": 825}, {"loss": 0.174755859375, "grad_norm": 0.4956880807876587, "learning_rate": 8.508880320952614e-06, "tIoU": 0.49635658264160154, "tRecall_50": 0.5375, "tRecall_75": 0.33125, "tRecall_90": 0.21875, "token_acc": 0.9361353711790393, "epoch": 0.27497101209209873, "step": 830}, {"loss": 0.1855224609375, "grad_norm": 1.0761417150497437, "learning_rate": 8.489720773831717e-06, "tIoU": 0.44313597679138184, "tRecall_50": 0.46875, "tRecall_75": 0.2875, "tRecall_90": 0.1875, "token_acc": 0.9296276013143483, "epoch": 0.2766274639721716, "step": 835}, {"loss": 0.19713134765625, "grad_norm": 0.7068014740943909, "learning_rate": 8.470460791210713e-06, "tIoU": 0.44738025665283204, "tRecall_50": 0.48125, "tRecall_75": 0.26875, "tRecall_90": 0.14375, "token_acc": 0.9266304347826086, "epoch": 0.2782839158522445, "step": 840}, {"loss": 0.1806640625, "grad_norm": 0.5590986013412476, "learning_rate": 8.451100927399272e-06, "tIoU": 0.4611639499664307, "tRecall_50": 0.51875, "tRecall_75": 0.325, "tRecall_90": 0.175, "token_acc": 0.9299206131946346, "epoch": 0.2799403677323174, "step": 845}, {"loss": 0.18641357421875, "grad_norm": 0.7248173356056213, "learning_rate": 8.431641739581688e-06, "tIoU": 0.4512800693511963, "tRecall_50": 0.48125, "tRecall_75": 0.30625, "tRecall_90": 0.18125, "token_acc": 0.9287849959228051, "epoch": 0.28159681961239025, "step": 850}, {"loss": 0.194873046875, "grad_norm": 0.4190671443939209, "learning_rate": 8.41208378780084e-06, "tIoU": 0.4547574520111084, "tRecall_50": 0.5, "tRecall_75": 0.28125, "tRecall_90": 0.19375, "token_acc": 0.9293286219081273, "epoch": 0.28325327149246315, "step": 855}, {"loss": 0.179071044921875, "grad_norm": 0.5794289708137512, "learning_rate": 8.392427634942069e-06, "tIoU": 0.5311003684997558, "tRecall_50": 0.58125, "tRecall_75": 0.35, "tRecall_90": 0.23125, "token_acc": 0.9356948228882834, "epoch": 0.284909723372536, "step": 860}, {"loss": 0.19573974609375, "grad_norm": 0.602472722530365, "learning_rate": 8.372673846716977e-06, "tIoU": 0.4420654296875, "tRecall_50": 0.46875, "tRecall_75": 0.26875, "tRecall_90": 0.175, "token_acc": 0.9296619411123228, "epoch": 0.2865661752526089, "step": 865}, {"loss": 0.20216064453125, "grad_norm": 0.8868810534477234, "learning_rate": 8.352822991647156e-06, "tIoU": 0.37321641445159914, "tRecall_50": 0.39375, "tRecall_75": 0.25, "tRecall_90": 0.15, "token_acc": 0.9270233196159122, "epoch": 0.2882226271326818, "step": 870}, {"loss": 0.185546875, "grad_norm": 0.5824978947639465, "learning_rate": 8.332875641047817e-06, "tIoU": 0.42017297744750975, "tRecall_50": 0.45625, "tRecall_75": 0.2875, "tRecall_90": 0.15, "token_acc": 0.930733569675484, "epoch": 0.28987907901275467, "step": 875}, {"loss": 0.18995361328125, "grad_norm": 0.5793182849884033, "learning_rate": 8.312832369011345e-06, "tIoU": 0.48091964721679686, "tRecall_50": 0.49375, "tRecall_75": 0.35625, "tRecall_90": 0.25625, "token_acc": 0.9309031556039173, "epoch": 0.2915355308928276, "step": 880}, {"loss": 0.1759521484375, "grad_norm": 0.44487687945365906, "learning_rate": 8.292693752390787e-06, "tIoU": 0.508469533920288, "tRecall_50": 0.54375, "tRecall_75": 0.325, "tRecall_90": 0.20625, "token_acc": 0.9326738114866722, "epoch": 0.29319198277290043, "step": 885}, {"loss": 0.1759033203125, "grad_norm": 1.132308006286621, "learning_rate": 8.272460370783238e-06, "tIoU": 0.44657158851623535, "tRecall_50": 0.475, "tRecall_75": 0.3, "tRecall_90": 0.175, "token_acc": 0.9313246471226927, "epoch": 0.29484843465297333, "step": 890}, {"loss": 0.17841796875, "grad_norm": 0.5606683492660522, "learning_rate": 8.25213280651317e-06, "tIoU": 0.45934696197509767, "tRecall_50": 0.50625, "tRecall_75": 0.3, "tRecall_90": 0.20625, "token_acc": 0.9355537503384782, "epoch": 0.29650488653304624, "step": 895}, {"loss": 0.19599609375, "grad_norm": 0.9272470474243164, "learning_rate": 8.231711644615667e-06, "tIoU": 0.3931166172027588, "tRecall_50": 0.4125, "tRecall_75": 0.2375, "tRecall_90": 0.14375, "token_acc": 0.9235374149659864, "epoch": 0.2981613384131191, "step": 900}, {"loss": 0.18623046875, "grad_norm": 0.586313784122467, "learning_rate": 8.211197472819584e-06, "tIoU": 0.46013169288635253, "tRecall_50": 0.475, "tRecall_75": 0.30625, "tRecall_90": 0.19375, "token_acc": 0.9299206131946346, "epoch": 0.299817790293192, "step": 905}, {"loss": 0.19263916015625, "grad_norm": 0.7881156802177429, "learning_rate": 8.190590881530644e-06, "tIoU": 0.4399313449859619, "tRecall_50": 0.5125, "tRecall_75": 0.3, "tRecall_90": 0.15625, "token_acc": 0.9282191780821918, "epoch": 0.30147424217326485, "step": 910}, {"loss": 0.17945556640625, "grad_norm": 0.6620801687240601, "learning_rate": 8.169892463814435e-06, "tIoU": 0.5031002521514892, "tRecall_50": 0.55, "tRecall_75": 0.34375, "tRecall_90": 0.20625, "token_acc": 0.9306122448979591, "epoch": 0.30313069405333776, "step": 915}, {"loss": 0.169970703125, "grad_norm": 0.5018839836120605, "learning_rate": 8.149102815379341e-06, "tIoU": 0.5297323703765869, "tRecall_50": 0.58125, "tRecall_75": 0.3875, "tRecall_90": 0.225, "token_acc": 0.9386252045826514, "epoch": 0.3047871459334106, "step": 920}, {"loss": 0.19112548828125, "grad_norm": 0.7223777174949646, "learning_rate": 8.128222534559406e-06, "tIoU": 0.49451518058776855, "tRecall_50": 0.525, "tRecall_75": 0.35, "tRecall_90": 0.225, "token_acc": 0.9320467518347377, "epoch": 0.3064435978134835, "step": 925}, {"loss": 0.184698486328125, "grad_norm": 0.5758301615715027, "learning_rate": 8.107252222297109e-06, "tIoU": 0.46448659896850586, "tRecall_50": 0.5, "tRecall_75": 0.325, "tRecall_90": 0.2, "token_acc": 0.9319672131147541, "epoch": 0.3081000496935564, "step": 930}, {"loss": 0.1820556640625, "grad_norm": 0.524838924407959, "learning_rate": 8.086192482126063e-06, "tIoU": 0.4440475940704346, "tRecall_50": 0.45625, "tRecall_75": 0.3125, "tRecall_90": 0.1625, "token_acc": 0.9292324442025041, "epoch": 0.3097565015736293, "step": 935}, {"loss": 0.1949951171875, "grad_norm": 0.563029944896698, "learning_rate": 8.065043920153658e-06, "tIoU": 0.4718454837799072, "tRecall_50": 0.4875, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.9278518921862238, "epoch": 0.3114129534537022, "step": 940}, {"loss": 0.19881591796875, "grad_norm": 0.6538600325584412, "learning_rate": 8.043807145043604e-06, "tIoU": 0.434765100479126, "tRecall_50": 0.48125, "tRecall_75": 0.3, "tRecall_90": 0.1625, "token_acc": 0.9281241491968418, "epoch": 0.31306940533377503, "step": 945}, {"loss": 0.20108642578125, "grad_norm": 0.47911351919174194, "learning_rate": 8.02248276799842e-06, "tIoU": 0.436666202545166, "tRecall_50": 0.4375, "tRecall_75": 0.28125, "tRecall_90": 0.18125, "token_acc": 0.9261580381471389, "epoch": 0.31472585721384794, "step": 950}, {"loss": 0.194287109375, "grad_norm": 0.8824852108955383, "learning_rate": 8.001071402741843e-06, "tIoU": 0.4746551990509033, "tRecall_50": 0.5, "tRecall_75": 0.35, "tRecall_90": 0.19375, "token_acc": 0.9284739284739285, "epoch": 0.31638230909392084, "step": 955}, {"loss": 0.1749969482421875, "grad_norm": 1.3164430856704712, "learning_rate": 7.979573665501161e-06, "tIoU": 0.45885448455810546, "tRecall_50": 0.46875, "tRecall_75": 0.34375, "tRecall_90": 0.2, "token_acc": 0.9300603400987384, "epoch": 0.3180387609739937, "step": 960}, {"loss": 0.18673095703125, "grad_norm": 0.744648814201355, "learning_rate": 7.957990174989487e-06, "tIoU": 0.4387523174285889, "tRecall_50": 0.45, "tRecall_75": 0.275, "tRecall_90": 0.175, "token_acc": 0.9319912948857454, "epoch": 0.3196952128540666, "step": 965}, {"loss": 0.18834228515625, "grad_norm": 0.4059516191482544, "learning_rate": 7.936321552387942e-06, "tIoU": 0.4600379467010498, "tRecall_50": 0.48125, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.9312789746386693, "epoch": 0.32135166473413945, "step": 970}, {"loss": 0.16092529296875, "grad_norm": 0.582944929599762, "learning_rate": 7.914568421327782e-06, "tIoU": 0.5187951564788819, "tRecall_50": 0.575, "tRecall_75": 0.35625, "tRecall_90": 0.2375, "token_acc": 0.9401475813063679, "epoch": 0.32300811661421236, "step": 975}, {"loss": 0.18634033203125, "grad_norm": 0.5389725565910339, "learning_rate": 7.892731407872446e-06, "tIoU": 0.45755715370178224, "tRecall_50": 0.45625, "tRecall_75": 0.31875, "tRecall_90": 0.175, "token_acc": 0.932322685435842, "epoch": 0.32466456849428527, "step": 980}, {"loss": 0.17401123046875, "grad_norm": 0.4865098297595978, "learning_rate": 7.870811140499543e-06, "tIoU": 0.5109381675720215, "tRecall_50": 0.5625, "tRecall_75": 0.3625, "tRecall_90": 0.2375, "token_acc": 0.9346049046321526, "epoch": 0.3263210203743581, "step": 985}, {"loss": 0.16990966796875, "grad_norm": 1.0831506252288818, "learning_rate": 7.848808250082763e-06, "tIoU": 0.48476033210754393, "tRecall_50": 0.5, "tRecall_75": 0.36875, "tRecall_90": 0.21875, "token_acc": 0.9362865736942849, "epoch": 0.327977472254431, "step": 990}, {"loss": 0.18963623046875, "grad_norm": 0.8786563277244568, "learning_rate": 7.826723369873715e-06, "tIoU": 0.4693743705749512, "tRecall_50": 0.5125, "tRecall_75": 0.3875, "tRecall_90": 0.21875, "token_acc": 0.9316031359826981, "epoch": 0.3296339241345039, "step": 995}, {"loss": 0.1913330078125, "grad_norm": 0.5306078195571899, "learning_rate": 7.80455713548371e-06, "tIoU": 0.41149063110351564, "tRecall_50": 0.44375, "tRecall_75": 0.21875, "tRecall_90": 0.13125, "token_acc": 0.9273073781649878, "epoch": 0.3312903760145768, "step": 1000}, {"loss": 0.19432373046875, "grad_norm": 0.5666636228561401, "learning_rate": 7.782310184865463e-06, "tIoU": 0.4761760711669922, "tRecall_50": 0.51875, "tRecall_75": 0.3375, "tRecall_90": 0.19375, "token_acc": 0.9221200648999459, "epoch": 0.33294682789464963, "step": 1005}, {"loss": 0.17784423828125, "grad_norm": 0.6108797788619995, "learning_rate": 7.759983158294727e-06, "tIoU": 0.4799777984619141, "tRecall_50": 0.5, "tRecall_75": 0.35625, "tRecall_90": 0.2125, "token_acc": 0.931607629427793, "epoch": 0.33460327977472254, "step": 1010}, {"loss": 0.18638916015625, "grad_norm": 0.540955126285553, "learning_rate": 7.737576698351878e-06, "tIoU": 0.4404765129089355, "tRecall_50": 0.48125, "tRecall_75": 0.26875, "tRecall_90": 0.125, "token_acc": 0.928024083196497, "epoch": 0.33625973165479545, "step": 1015}, {"loss": 0.1887939453125, "grad_norm": 0.5134669542312622, "learning_rate": 7.715091449903414e-06, "tIoU": 0.4650413990020752, "tRecall_50": 0.5, "tRecall_75": 0.3, "tRecall_90": 0.1375, "token_acc": 0.9318986652138382, "epoch": 0.3379161835348683, "step": 1020}, {"loss": 0.18924560546875, "grad_norm": 1.0283092260360718, "learning_rate": 7.692528060083392e-06, "tIoU": 0.4443990230560303, "tRecall_50": 0.4625, "tRecall_75": 0.28125, "tRecall_90": 0.20625, "token_acc": 0.9303849303849304, "epoch": 0.3395726354149412, "step": 1025}, {"loss": 0.17459716796875, "grad_norm": 0.5505083799362183, "learning_rate": 7.669887178274815e-06, "tIoU": 0.5016412734985352, "tRecall_50": 0.50625, "tRecall_75": 0.425, "tRecall_90": 0.275, "token_acc": 0.9353260869565218, "epoch": 0.34122908729501406, "step": 1030}, {"loss": 0.177587890625, "grad_norm": 0.6912838816642761, "learning_rate": 7.647169456090925e-06, "tIoU": 0.48855109214782716, "tRecall_50": 0.56875, "tRecall_75": 0.34375, "tRecall_90": 0.21875, "token_acc": 0.935685210312076, "epoch": 0.34288553917508696, "step": 1035}, {"loss": 0.17293701171875, "grad_norm": 0.778060257434845, "learning_rate": 7.624375547356473e-06, "tIoU": 0.49210290908813475, "tRecall_50": 0.50625, "tRecall_75": 0.375, "tRecall_90": 0.23125, "token_acc": 0.936542669584245, "epoch": 0.34454199105515987, "step": 1040}, {"loss": 0.178436279296875, "grad_norm": 0.9876009225845337, "learning_rate": 7.601506108088874e-06, "tIoU": 0.5314978122711181, "tRecall_50": 0.575, "tRecall_75": 0.3875, "tRecall_90": 0.2375, "token_acc": 0.9381639880141651, "epoch": 0.3461984429352327, "step": 1045}, {"loss": 0.18966064453125, "grad_norm": 0.8202823400497437, "learning_rate": 7.578561796479347e-06, "tIoU": 0.4976532459259033, "tRecall_50": 0.5375, "tRecall_75": 0.325, "tRecall_90": 0.175, "token_acc": 0.9336076817558299, "epoch": 0.3478548948153056, "step": 1050}, {"loss": 0.19049072265625, "grad_norm": 0.7027230858802795, "learning_rate": 7.55554327287397e-06, "tIoU": 0.4707040786743164, "tRecall_50": 0.5, "tRecall_75": 0.31875, "tRecall_90": 0.16875, "token_acc": 0.9283964062074599, "epoch": 0.3495113466953785, "step": 1055}, {"loss": 0.1842529296875, "grad_norm": 0.5952333211898804, "learning_rate": 7.532451199754662e-06, "tIoU": 0.5045681476593018, "tRecall_50": 0.55, "tRecall_75": 0.3625, "tRecall_90": 0.2125, "token_acc": 0.9287849959228051, "epoch": 0.3511677985754514, "step": 1060}, {"loss": 0.178466796875, "grad_norm": 0.6814955472946167, "learning_rate": 7.509286241720133e-06, "tIoU": 0.5122104167938233, "tRecall_50": 0.5875, "tRecall_75": 0.34375, "tRecall_90": 0.19375, "token_acc": 0.9327525183773482, "epoch": 0.3528242504555243, "step": 1065}, {"loss": 0.18367919921875, "grad_norm": 0.7592156529426575, "learning_rate": 7.486049065466743e-06, "tIoU": 0.4885748863220215, "tRecall_50": 0.525, "tRecall_75": 0.34375, "tRecall_90": 0.23125, "token_acc": 0.935659760087241, "epoch": 0.35448070233559714, "step": 1070}, {"loss": 0.18165283203125, "grad_norm": 0.6961227655410767, "learning_rate": 7.462740339769323e-06, "tIoU": 0.46857433319091796, "tRecall_50": 0.5, "tRecall_75": 0.3125, "tRecall_90": 0.18125, "token_acc": 0.9313725490196079, "epoch": 0.35613715421567005, "step": 1075}, {"loss": 0.18089599609375, "grad_norm": 0.5508667230606079, "learning_rate": 7.439360735461923e-06, "tIoU": 0.48640642166137693, "tRecall_50": 0.49375, "tRecall_75": 0.36875, "tRecall_90": 0.225, "token_acc": 0.9371241115363587, "epoch": 0.3577936060957429, "step": 1080}, {"loss": 0.1744140625, "grad_norm": 0.5022640824317932, "learning_rate": 7.415910925418508e-06, "tIoU": 0.5186983108520508, "tRecall_50": 0.525, "tRecall_75": 0.375, "tRecall_90": 0.25, "token_acc": 0.9383356070941337, "epoch": 0.3594500579758158, "step": 1085}, {"loss": 0.1818603515625, "grad_norm": 0.5923255085945129, "learning_rate": 7.392391584533591e-06, "tIoU": 0.47256951332092284, "tRecall_50": 0.51875, "tRecall_75": 0.33125, "tRecall_90": 0.175, "token_acc": 0.9305177111716622, "epoch": 0.3611065098558887, "step": 1090}, {"loss": 0.1818603515625, "grad_norm": 0.5832929611206055, "learning_rate": 7.368803389702805e-06, "tIoU": 0.49889507293701174, "tRecall_50": 0.55625, "tRecall_75": 0.325, "tRecall_90": 0.2, "token_acc": 0.9336986301369863, "epoch": 0.36276296173596156, "step": 1095}, {"loss": 0.1832275390625, "grad_norm": 0.5508763790130615, "learning_rate": 7.345147019803433e-06, "tIoU": 0.47658720016479494, "tRecall_50": 0.50625, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.9340958605664488, "epoch": 0.36441941361603447, "step": 1100}, {"loss": 0.1625244140625, "grad_norm": 0.49825698137283325, "learning_rate": 7.321423155674858e-06, "tIoU": 0.5512962341308594, "tRecall_50": 0.5875, "tRecall_75": 0.41875, "tRecall_90": 0.26875, "token_acc": 0.940515350877193, "epoch": 0.3660758654961073, "step": 1105}, {"loss": 0.1703369140625, "grad_norm": 0.4958120286464691, "learning_rate": 7.297632480098973e-06, "tIoU": 0.529345703125, "tRecall_50": 0.55, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.9389084985066521, "epoch": 0.36773231737618023, "step": 1110}, {"loss": 0.17938232421875, "grad_norm": 0.5963060259819031, "learning_rate": 7.27377567778053e-06, "tIoU": 0.5171398639678955, "tRecall_50": 0.575, "tRecall_75": 0.375, "tRecall_90": 0.1875, "token_acc": 0.9303523035230352, "epoch": 0.3693887692562531, "step": 1115}, {"loss": 0.19072265625, "grad_norm": 0.7840315103530884, "learning_rate": 7.249853435327435e-06, "tIoU": 0.45734376907348634, "tRecall_50": 0.46875, "tRecall_75": 0.29375, "tRecall_90": 0.16875, "token_acc": 0.9290375203915171, "epoch": 0.371045221136326, "step": 1120}, {"loss": 0.18948974609375, "grad_norm": 0.6449282169342041, "learning_rate": 7.225866441230989e-06, "tIoU": 0.4800254821777344, "tRecall_50": 0.50625, "tRecall_75": 0.325, "tRecall_90": 0.2, "token_acc": 0.9292709466811752, "epoch": 0.3727016730163989, "step": 1125}, {"loss": 0.17431640625, "grad_norm": 0.4980352520942688, "learning_rate": 7.201815385846062e-06, "tIoU": 0.5075998783111573, "tRecall_50": 0.53125, "tRecall_75": 0.34375, "tRecall_90": 0.25625, "token_acc": 0.9373461012311901, "epoch": 0.37435812489647174, "step": 1130}, {"loss": 0.186663818359375, "grad_norm": 1.3992658853530884, "learning_rate": 7.177700961371239e-06, "tIoU": 0.46782970428466797, "tRecall_50": 0.475, "tRecall_75": 0.3375, "tRecall_90": 0.20625, "token_acc": 0.9325751421608448, "epoch": 0.37601457677654465, "step": 1135}, {"loss": 0.19114990234375, "grad_norm": 0.5823772549629211, "learning_rate": 7.1535238618288915e-06, "tIoU": 0.4538295269012451, "tRecall_50": 0.46875, "tRecall_75": 0.31875, "tRecall_90": 0.20625, "token_acc": 0.9287069666576308, "epoch": 0.3776710286566175, "step": 1140}, {"loss": 0.17542724609375, "grad_norm": 0.5577098727226257, "learning_rate": 7.1292847830452e-06, "tIoU": 0.45658197402954104, "tRecall_50": 0.49375, "tRecall_75": 0.325, "tRecall_90": 0.2125, "token_acc": 0.9328621908127208, "epoch": 0.3793274805366904, "step": 1145}, {"loss": 0.1681640625, "grad_norm": 0.5321166515350342, "learning_rate": 7.104984422630134e-06, "tIoU": 0.541117000579834, "tRecall_50": 0.61875, "tRecall_75": 0.35, "tRecall_90": 0.2125, "token_acc": 0.9354309165526675, "epoch": 0.3809839324167633, "step": 1150}, {"loss": 0.173046875, "grad_norm": 0.5487158894538879, "learning_rate": 7.080623479957372e-06, "tIoU": 0.5375588893890381, "tRecall_50": 0.5875, "tRecall_75": 0.36875, "tRecall_90": 0.25, "token_acc": 0.937857726901063, "epoch": 0.38264038429683617, "step": 1155}, {"loss": 0.18504638671875, "grad_norm": 0.640888512134552, "learning_rate": 7.056202656144173e-06, "tIoU": 0.4445330142974854, "tRecall_50": 0.4625, "tRecall_75": 0.28125, "tRecall_90": 0.16875, "token_acc": 0.930327868852459, "epoch": 0.3842968361769091, "step": 1160}, {"loss": 0.1691162109375, "grad_norm": 0.47475653886795044, "learning_rate": 7.031722654031192e-06, "tIoU": 0.4704757213592529, "tRecall_50": 0.50625, "tRecall_75": 0.35, "tRecall_90": 0.2, "token_acc": 0.9352733206418276, "epoch": 0.3859532880569819, "step": 1165}, {"loss": 0.19022216796875, "grad_norm": 0.7627384662628174, "learning_rate": 7.007184178162268e-06, "tIoU": 0.48237972259521483, "tRecall_50": 0.51875, "tRecall_75": 0.31875, "tRecall_90": 0.2, "token_acc": 0.9302832244008714, "epoch": 0.38760973993705483, "step": 1170}, {"loss": 0.17056884765625, "grad_norm": 0.7071655988693237, "learning_rate": 6.982587934764133e-06, "tIoU": 0.5011544227600098, "tRecall_50": 0.55, "tRecall_75": 0.325, "tRecall_90": 0.2, "token_acc": 0.9356677524429967, "epoch": 0.38926619181712774, "step": 1175}, {"loss": 0.1821044921875, "grad_norm": 0.6485012173652649, "learning_rate": 6.9579346317260875e-06, "tIoU": 0.5037876605987549, "tRecall_50": 0.5625, "tRecall_75": 0.3625, "tRecall_90": 0.23125, "token_acc": 0.9321340964840557, "epoch": 0.3909226436972006, "step": 1180}, {"loss": 0.167529296875, "grad_norm": 0.6403648853302002, "learning_rate": 6.933224978579632e-06, "tIoU": 0.5066840648651123, "tRecall_50": 0.55, "tRecall_75": 0.3875, "tRecall_90": 0.2125, "token_acc": 0.938891906572515, "epoch": 0.3925790955772735, "step": 1185}, {"loss": 0.1739013671875, "grad_norm": 0.5812480449676514, "learning_rate": 6.90845968647805e-06, "tIoU": 0.46291217803955076, "tRecall_50": 0.525, "tRecall_75": 0.33125, "tRecall_90": 0.18125, "token_acc": 0.9338415464198203, "epoch": 0.39423554745734635, "step": 1190}, {"loss": 0.17183837890625, "grad_norm": 0.526985764503479, "learning_rate": 6.883639468175926e-06, "tIoU": 0.4848320484161377, "tRecall_50": 0.55625, "tRecall_75": 0.35, "tRecall_90": 0.2, "token_acc": 0.9363213992894234, "epoch": 0.39589199933741925, "step": 1195}, {"loss": 0.1871826171875, "grad_norm": 0.7262851595878601, "learning_rate": 6.85876503800865e-06, "tIoU": 0.4523651599884033, "tRecall_50": 0.4875, "tRecall_75": 0.3125, "tRecall_90": 0.2125, "token_acc": 0.9293066088840737, "epoch": 0.3975484512174921, "step": 1200}, {"loss": 0.19154052734375, "grad_norm": 0.5683233141899109, "learning_rate": 6.833837111871845e-06, "tIoU": 0.43550262451171873, "tRecall_50": 0.425, "tRecall_75": 0.3125, "tRecall_90": 0.21875, "token_acc": 0.9319357473454941, "epoch": 0.399204903097565, "step": 1205}, {"loss": 0.177423095703125, "grad_norm": 0.703508734703064, "learning_rate": 6.808856407200772e-06, "tIoU": 0.525466012954712, "tRecall_50": 0.58125, "tRecall_75": 0.3625, "tRecall_90": 0.2125, "token_acc": 0.9351248642779587, "epoch": 0.4008613549776379, "step": 1210}, {"loss": 0.16162109375, "grad_norm": 0.5343628525733948, "learning_rate": 6.7838236429496766e-06, "tIoU": 0.5424517631530762, "tRecall_50": 0.55625, "tRecall_75": 0.425, "tRecall_90": 0.29375, "token_acc": 0.9389645776566757, "epoch": 0.40251780685771077, "step": 1215}, {"loss": 0.1860107421875, "grad_norm": 0.3772442936897278, "learning_rate": 6.758739539571096e-06, "tIoU": 0.4494819641113281, "tRecall_50": 0.44375, "tRecall_75": 0.3, "tRecall_90": 0.19375, "token_acc": 0.928922908693275, "epoch": 0.4041742587377837, "step": 1220}, {"loss": 0.17764892578125, "grad_norm": 0.4595545530319214, "learning_rate": 6.733604818995133e-06, "tIoU": 0.5240142822265625, "tRecall_50": 0.59375, "tRecall_75": 0.3625, "tRecall_90": 0.18125, "token_acc": 0.9381639880141651, "epoch": 0.4058307106178565, "step": 1225}, {"loss": 0.1759033203125, "grad_norm": 0.7019212245941162, "learning_rate": 6.708420204608671e-06, "tIoU": 0.43599753379821776, "tRecall_50": 0.4625, "tRecall_75": 0.2875, "tRecall_90": 0.14375, "token_acc": 0.9295238095238095, "epoch": 0.40748716249792943, "step": 1230}, {"loss": 0.153509521484375, "grad_norm": 0.5394700169563293, "learning_rate": 6.683186421234552e-06, "tIoU": 0.5693307876586914, "tRecall_50": 0.5875, "tRecall_75": 0.43125, "tRecall_90": 0.29375, "token_acc": 0.9454693434617472, "epoch": 0.40914361437800234, "step": 1235}, {"loss": 0.182177734375, "grad_norm": 0.9691959619522095, "learning_rate": 6.657904195110727e-06, "tIoU": 0.4553550720214844, "tRecall_50": 0.48125, "tRecall_75": 0.3, "tRecall_90": 0.18125, "token_acc": 0.9309120699071546, "epoch": 0.4108000662580752, "step": 1240}, {"loss": 0.1696044921875, "grad_norm": 0.9197001457214355, "learning_rate": 6.6325742538693424e-06, "tIoU": 0.5319690227508544, "tRecall_50": 0.5875, "tRecall_75": 0.4, "tRecall_90": 0.225, "token_acc": 0.9379084967320261, "epoch": 0.4124565181381481, "step": 1245}, {"loss": 0.1841552734375, "grad_norm": 0.7810382843017578, "learning_rate": 6.607197326515808e-06, "tIoU": 0.49736499786376953, "tRecall_50": 0.5125, "tRecall_75": 0.35, "tRecall_90": 0.2625, "token_acc": 0.9347237269772481, "epoch": 0.41411297001822095, "step": 1250}, {"loss": 0.1776611328125, "grad_norm": 0.5846786499023438, "learning_rate": 6.58177414340781e-06, "tIoU": 0.495667028427124, "tRecall_50": 0.50625, "tRecall_75": 0.36875, "tRecall_90": 0.25, "token_acc": 0.9368852459016394, "epoch": 0.41576942189829386, "step": 1255}, {"loss": 0.17413330078125, "grad_norm": 0.5197479724884033, "learning_rate": 6.556305436234292e-06, "tIoU": 0.5085415363311767, "tRecall_50": 0.53125, "tRecall_75": 0.36875, "tRecall_90": 0.20625, "token_acc": 0.9321477428180575, "epoch": 0.41742587377836676, "step": 1260}, {"loss": 0.1685791015625, "grad_norm": 0.4828190207481384, "learning_rate": 6.530791937994397e-06, "tIoU": 0.557618522644043, "tRecall_50": 0.61875, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.9379254015790907, "epoch": 0.4190823256584396, "step": 1265}, {"loss": 0.18017578125, "grad_norm": 0.5395064949989319, "learning_rate": 6.505234382976375e-06, "tIoU": 0.49621214866638186, "tRecall_50": 0.54375, "tRecall_75": 0.325, "tRecall_90": 0.23125, "token_acc": 0.935562805872757, "epoch": 0.4207387775385125, "step": 1270}, {"loss": 0.166064453125, "grad_norm": 0.6637852787971497, "learning_rate": 6.479633506736447e-06, "tIoU": 0.4753843307495117, "tRecall_50": 0.4875, "tRecall_75": 0.38125, "tRecall_90": 0.23125, "token_acc": 0.9365600218758545, "epoch": 0.42239522941858537, "step": 1275}, {"loss": 0.15682373046875, "grad_norm": 0.5723521113395691, "learning_rate": 6.453990046077634e-06, "tIoU": 0.5220492362976075, "tRecall_50": 0.575, "tRecall_75": 0.36875, "tRecall_90": 0.2375, "token_acc": 0.9406941787373599, "epoch": 0.4240516812986583, "step": 1280}, {"loss": 0.1759033203125, "grad_norm": 0.5472649335861206, "learning_rate": 6.42830473902855e-06, "tIoU": 0.4785460472106934, "tRecall_50": 0.5, "tRecall_75": 0.3125, "tRecall_90": 0.2375, "token_acc": 0.937720968180582, "epoch": 0.4257081331787312, "step": 1285}, {"loss": 0.18299560546875, "grad_norm": 0.5641521215438843, "learning_rate": 6.402578324822169e-06, "tIoU": 0.4435619354248047, "tRecall_50": 0.4625, "tRecall_75": 0.26875, "tRecall_90": 0.18125, "token_acc": 0.9317621266100301, "epoch": 0.42736458505880404, "step": 1290}, {"loss": 0.184521484375, "grad_norm": 0.4343439042568207, "learning_rate": 6.3768115438745434e-06, "tIoU": 0.500169038772583, "tRecall_50": 0.54375, "tRecall_75": 0.3375, "tRecall_90": 0.20625, "token_acc": 0.9354926510615134, "epoch": 0.42902103693887694, "step": 1295}, {"loss": 0.1848388671875, "grad_norm": 0.6340326070785522, "learning_rate": 6.351005137763494e-06, "tIoU": 0.513594675064087, "tRecall_50": 0.5875, "tRecall_75": 0.3375, "tRecall_90": 0.175, "token_acc": 0.9330430048992924, "epoch": 0.4306774888189498, "step": 1300}, {"loss": 0.18204345703125, "grad_norm": 0.6288878917694092, "learning_rate": 6.3251598492072685e-06, "tIoU": 0.44141411781311035, "tRecall_50": 0.45, "tRecall_75": 0.28125, "tRecall_90": 0.175, "token_acc": 0.9306200491668943, "epoch": 0.4323339406990227, "step": 1305}, {"loss": 0.16866455078125, "grad_norm": 0.7356960773468018, "learning_rate": 6.299276422043171e-06, "tIoU": 0.4666309833526611, "tRecall_50": 0.525, "tRecall_75": 0.35, "tRecall_90": 0.23125, "token_acc": 0.9368852459016394, "epoch": 0.43399039257909555, "step": 1310}, {"loss": 0.1773681640625, "grad_norm": 0.6025828123092651, "learning_rate": 6.273355601206143e-06, "tIoU": 0.48667373657226565, "tRecall_50": 0.54375, "tRecall_75": 0.31875, "tRecall_90": 0.2, "token_acc": 0.9369171695008228, "epoch": 0.43564684445916846, "step": 1315}, {"loss": 0.180078125, "grad_norm": 0.4284733831882477, "learning_rate": 6.247398132707339e-06, "tIoU": 0.49059696197509767, "tRecall_50": 0.5125, "tRecall_75": 0.36875, "tRecall_90": 0.23125, "token_acc": 0.9323737099402498, "epoch": 0.43730329633924137, "step": 1320}, {"loss": 0.18408203125, "grad_norm": 0.5862426161766052, "learning_rate": 6.221404763612635e-06, "tIoU": 0.4608908653259277, "tRecall_50": 0.5125, "tRecall_75": 0.3, "tRecall_90": 0.18125, "token_acc": 0.9308365226899945, "epoch": 0.4389597482193142, "step": 1325}, {"loss": 0.1912109375, "grad_norm": 1.026079773902893, "learning_rate": 6.1953762420211504e-06, "tIoU": 0.489334774017334, "tRecall_50": 0.5125, "tRecall_75": 0.3625, "tRecall_90": 0.18125, "token_acc": 0.9306122448979591, "epoch": 0.4406162000993871, "step": 1330}, {"loss": 0.172412109375, "grad_norm": 0.6226865649223328, "learning_rate": 6.169313317043702e-06, "tIoU": 0.482784366607666, "tRecall_50": 0.525, "tRecall_75": 0.31875, "tRecall_90": 0.18125, "token_acc": 0.9340809628008753, "epoch": 0.44227265197946, "step": 1335}, {"loss": 0.1625244140625, "grad_norm": 1.1509168148040771, "learning_rate": 6.143216738781251e-06, "tIoU": 0.5060029983520508, "tRecall_50": 0.55625, "tRecall_75": 0.35, "tRecall_90": 0.2125, "token_acc": 0.9401475813063679, "epoch": 0.4439291038595329, "step": 1340}, {"loss": 0.171435546875, "grad_norm": 0.6278771162033081, "learning_rate": 6.117087258303314e-06, "tIoU": 0.4706563949584961, "tRecall_50": 0.48125, "tRecall_75": 0.375, "tRecall_90": 0.21875, "token_acc": 0.9342069342069342, "epoch": 0.4455855557396058, "step": 1345}, {"loss": 0.182470703125, "grad_norm": 0.4377037286758423, "learning_rate": 6.09092562762634e-06, "tIoU": 0.43801226615905764, "tRecall_50": 0.45, "tRecall_75": 0.3375, "tRecall_90": 0.2125, "token_acc": 0.9311475409836065, "epoch": 0.44724200761967864, "step": 1350}, {"loss": 0.1802001953125, "grad_norm": 0.8750120401382446, "learning_rate": 6.064732599692079e-06, "tIoU": 0.46314353942871095, "tRecall_50": 0.50625, "tRecall_75": 0.29375, "tRecall_90": 0.20625, "token_acc": 0.935025935025935, "epoch": 0.44889845949975155, "step": 1355}, {"loss": 0.176611328125, "grad_norm": 1.339943289756775, "learning_rate": 6.038508928345906e-06, "tIoU": 0.46857519149780275, "tRecall_50": 0.50625, "tRecall_75": 0.31875, "tRecall_90": 0.2, "token_acc": 0.9331514324693042, "epoch": 0.4505549113798244, "step": 1360}, {"loss": 0.16986083984375, "grad_norm": 0.38020825386047363, "learning_rate": 6.012255368315123e-06, "tIoU": 0.4757106781005859, "tRecall_50": 0.5, "tRecall_75": 0.35, "tRecall_90": 0.18125, "token_acc": 0.9373808875578546, "epoch": 0.4522113632598973, "step": 1365}, {"loss": 0.17435302734375, "grad_norm": 0.5417476296424866, "learning_rate": 5.985972675187239e-06, "tIoU": 0.4798302173614502, "tRecall_50": 0.5, "tRecall_75": 0.30625, "tRecall_90": 0.2, "token_acc": 0.9314285714285714, "epoch": 0.4538678151399702, "step": 1370}, {"loss": 0.186907958984375, "grad_norm": 0.6107211709022522, "learning_rate": 5.959661605388229e-06, "tIoU": 0.46599550247192384, "tRecall_50": 0.45, "tRecall_75": 0.3375, "tRecall_90": 0.21875, "token_acc": 0.930219929405376, "epoch": 0.45552426702004306, "step": 1375}, {"loss": 0.163482666015625, "grad_norm": 1.218456506729126, "learning_rate": 5.933322916160755e-06, "tIoU": 0.5246377944946289, "tRecall_50": 0.5625, "tRecall_75": 0.375, "tRecall_90": 0.23125, "token_acc": 0.9383524277141299, "epoch": 0.45718071890011597, "step": 1380}, {"loss": 0.17071533203125, "grad_norm": 0.46208956837654114, "learning_rate": 5.906957365542376e-06, "tIoU": 0.504871129989624, "tRecall_50": 0.54375, "tRecall_75": 0.3625, "tRecall_90": 0.2125, "token_acc": 0.9363686231486561, "epoch": 0.4588371707801888, "step": 1385}, {"loss": 0.18258056640625, "grad_norm": 0.44012129306793213, "learning_rate": 5.880565712343738e-06, "tIoU": 0.49883341789245605, "tRecall_50": 0.5375, "tRecall_75": 0.35, "tRecall_90": 0.19375, "token_acc": 0.934135009565455, "epoch": 0.4604936226602617, "step": 1390}, {"loss": 0.16441650390625, "grad_norm": 0.6960217356681824, "learning_rate": 5.854148716126722e-06, "tIoU": 0.5241262912750244, "tRecall_50": 0.55625, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.9384531590413944, "epoch": 0.4621500745403346, "step": 1395}, {"loss": 0.17698974609375, "grad_norm": 1.0893362760543823, "learning_rate": 5.827707137182597e-06, "tIoU": 0.4990966796875, "tRecall_50": 0.55, "tRecall_75": 0.375, "tRecall_90": 0.2, "token_acc": 0.9318243796018544, "epoch": 0.4638065264204075, "step": 1400}, {"loss": 0.166265869140625, "grad_norm": 0.6267316937446594, "learning_rate": 5.801241736510128e-06, "tIoU": 0.5334889411926269, "tRecall_50": 0.55625, "tRecall_75": 0.39375, "tRecall_90": 0.25625, "token_acc": 0.9387588459444747, "epoch": 0.4654629783004804, "step": 1405}, {"loss": 0.18515625, "grad_norm": 0.5143325924873352, "learning_rate": 5.774753275793682e-06, "tIoU": 0.4607995986938477, "tRecall_50": 0.5125, "tRecall_75": 0.29375, "tRecall_90": 0.175, "token_acc": 0.930958904109589, "epoch": 0.46711943018055324, "step": 1410}, {"loss": 0.16951904296875, "grad_norm": 0.4813113510608673, "learning_rate": 5.748242517381301e-06, "tIoU": 0.5051668643951416, "tRecall_50": 0.56875, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.9371086305472366, "epoch": 0.46877588206062615, "step": 1415}, {"loss": 0.160137939453125, "grad_norm": 0.7769686579704285, "learning_rate": 5.721710224262759e-06, "tIoU": 0.5131546974182128, "tRecall_50": 0.5625, "tRecall_75": 0.3875, "tRecall_90": 0.24375, "token_acc": 0.9359391965255157, "epoch": 0.470432333940699, "step": 1420}, {"loss": 0.18056640625, "grad_norm": 0.5152138471603394, "learning_rate": 5.695157160047618e-06, "tIoU": 0.4545309066772461, "tRecall_50": 0.4875, "tRecall_75": 0.325, "tRecall_90": 0.23125, "token_acc": 0.9355102040816327, "epoch": 0.4720887858207719, "step": 1425}, {"loss": 0.19742431640625, "grad_norm": 0.6680763959884644, "learning_rate": 5.66858408894323e-06, "tIoU": 0.41425528526306155, "tRecall_50": 0.43125, "tRecall_75": 0.25, "tRecall_90": 0.1625, "token_acc": 0.9210740439381611, "epoch": 0.4737452377008448, "step": 1430}, {"loss": 0.16871337890625, "grad_norm": 0.45692986249923706, "learning_rate": 5.641991775732756e-06, "tIoU": 0.5347061634063721, "tRecall_50": 0.58125, "tRecall_75": 0.4125, "tRecall_90": 0.26875, "token_acc": 0.9424069546318935, "epoch": 0.47540168958091766, "step": 1435}, {"loss": 0.15567626953125, "grad_norm": 0.6588415503501892, "learning_rate": 5.615380985753152e-06, "tIoU": 0.5204037189483642, "tRecall_50": 0.5625, "tRecall_75": 0.34375, "tRecall_90": 0.23125, "token_acc": 0.935984745301008, "epoch": 0.47705814146099057, "step": 1440}, {"loss": 0.16971435546875, "grad_norm": 0.6113463044166565, "learning_rate": 5.588752484873146e-06, "tIoU": 0.5141745567321777, "tRecall_50": 0.55625, "tRecall_75": 0.4, "tRecall_90": 0.225, "token_acc": 0.9385917030567685, "epoch": 0.4787145933410634, "step": 1445}, {"loss": 0.17086181640625, "grad_norm": 0.45739471912384033, "learning_rate": 5.562107039471187e-06, "tIoU": 0.5260482788085937, "tRecall_50": 0.5875, "tRecall_75": 0.3875, "tRecall_90": 0.2375, "token_acc": 0.9395438307227261, "epoch": 0.48037104522113633, "step": 1450}, {"loss": 0.17530517578125, "grad_norm": 0.510820209980011, "learning_rate": 5.535445416413396e-06, "tIoU": 0.51797456741333, "tRecall_50": 0.55625, "tRecall_75": 0.36875, "tRecall_90": 0.2375, "token_acc": 0.9394022484233616, "epoch": 0.48202749710120923, "step": 1455}, {"loss": 0.17376708984375, "grad_norm": 0.6514920592308044, "learning_rate": 5.508768383031497e-06, "tIoU": 0.5032854080200195, "tRecall_50": 0.55, "tRecall_75": 0.38125, "tRecall_90": 0.2125, "token_acc": 0.9332247557003257, "epoch": 0.4836839489812821, "step": 1460}, {"loss": 0.18035888671875, "grad_norm": 0.5166162252426147, "learning_rate": 5.482076707100723e-06, "tIoU": 0.5216215133666993, "tRecall_50": 0.5625, "tRecall_75": 0.375, "tRecall_90": 0.275, "token_acc": 0.935230352303523, "epoch": 0.485340400861355, "step": 1465}, {"loss": 0.16441650390625, "grad_norm": 0.4673994183540344, "learning_rate": 5.455371156817726e-06, "tIoU": 0.5096771717071533, "tRecall_50": 0.58125, "tRecall_75": 0.34375, "tRecall_90": 0.175, "token_acc": 0.9382850901146914, "epoch": 0.48699685274142784, "step": 1470}, {"loss": 0.1756103515625, "grad_norm": 0.5138466358184814, "learning_rate": 5.428652500778472e-06, "tIoU": 0.45171260833740234, "tRecall_50": 0.45625, "tRecall_75": 0.3, "tRecall_90": 0.175, "token_acc": 0.9321941958231624, "epoch": 0.48865330462150075, "step": 1475}, {"loss": 0.172747802734375, "grad_norm": 0.3812209963798523, "learning_rate": 5.401921507956113e-06, "tIoU": 0.509559154510498, "tRecall_50": 0.55, "tRecall_75": 0.375, "tRecall_90": 0.2375, "token_acc": 0.9372111986952977, "epoch": 0.49030975650157366, "step": 1480}, {"loss": 0.188433837890625, "grad_norm": 0.6065695285797119, "learning_rate": 5.375178947678854e-06, "tIoU": 0.44904017448425293, "tRecall_50": 0.45625, "tRecall_75": 0.3, "tRecall_90": 0.18125, "token_acc": 0.9303145336225597, "epoch": 0.4919662083816465, "step": 1485}, {"loss": 0.16478271484375, "grad_norm": 0.9850072264671326, "learning_rate": 5.348425589607823e-06, "tIoU": 0.5062573909759521, "tRecall_50": 0.54375, "tRecall_75": 0.36875, "tRecall_90": 0.24375, "token_acc": 0.9359255202628697, "epoch": 0.4936226602617194, "step": 1490}, {"loss": 0.17548828125, "grad_norm": 0.7354980707168579, "learning_rate": 5.321662203714909e-06, "tIoU": 0.5269069671630859, "tRecall_50": 0.58125, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.9357123399618632, "epoch": 0.49527911214179227, "step": 1495}, {"loss": 0.17823486328125, "grad_norm": 0.4693378806114197, "learning_rate": 5.294889560260605e-06, "tIoU": 0.4679423809051514, "tRecall_50": 0.50625, "tRecall_75": 0.30625, "tRecall_90": 0.19375, "token_acc": 0.9336971350613915, "epoch": 0.4969355640218652, "step": 1500}, {"loss": 0.166357421875, "grad_norm": 0.5371018648147583, "learning_rate": 5.268108429771841e-06, "tIoU": 0.490740442276001, "tRecall_50": 0.5625, "tRecall_75": 0.35625, "tRecall_90": 0.23125, "token_acc": 0.9367675115835378, "epoch": 0.498592015901938, "step": 1505}, {"loss": 0.17530517578125, "grad_norm": 0.6271441578865051, "learning_rate": 5.2413195830198075e-06, "tIoU": 0.48148627281188966, "tRecall_50": 0.5375, "tRecall_75": 0.34375, "tRecall_90": 0.2, "token_acc": 0.9345026034530008, "epoch": 0.5002484677820109, "step": 1510}, {"loss": 0.1841796875, "grad_norm": 0.7134257555007935, "learning_rate": 5.214523790997773e-06, "tIoU": 0.5060761451721192, "tRecall_50": 0.5125, "tRecall_75": 0.36875, "tRecall_90": 0.24375, "token_acc": 0.9378746594005449, "epoch": 0.5019049196620838, "step": 1515}, {"loss": 0.18895263671875, "grad_norm": 0.5665849447250366, "learning_rate": 5.187721824898889e-06, "tIoU": 0.48246068954467775, "tRecall_50": 0.50625, "tRecall_75": 0.34375, "tRecall_90": 0.19375, "token_acc": 0.9330634278002699, "epoch": 0.5035613715421567, "step": 1520}, {"loss": 0.17012939453125, "grad_norm": 0.7415385842323303, "learning_rate": 5.160914456094005e-06, "tIoU": 0.5014803409576416, "tRecall_50": 0.5375, "tRecall_75": 0.38125, "tRecall_90": 0.20625, "token_acc": 0.9352459016393443, "epoch": 0.5052178234222295, "step": 1525}, {"loss": 0.17379150390625, "grad_norm": 0.5564184188842773, "learning_rate": 5.134102456109461e-06, "tIoU": 0.46170501708984374, "tRecall_50": 0.48125, "tRecall_75": 0.31875, "tRecall_90": 0.19375, "token_acc": 0.9353342428376534, "epoch": 0.5068742753023024, "step": 1530}, {"loss": 0.16895751953125, "grad_norm": 0.5766486525535583, "learning_rate": 5.10728659660488e-06, "tIoU": 0.4611226558685303, "tRecall_50": 0.5375, "tRecall_75": 0.30625, "tRecall_90": 0.1625, "token_acc": 0.9382344902978956, "epoch": 0.5085307271823754, "step": 1535}, {"loss": 0.1735595703125, "grad_norm": 0.48594367504119873, "learning_rate": 5.080467649350968e-06, "tIoU": 0.47720837593078613, "tRecall_50": 0.51875, "tRecall_75": 0.33125, "tRecall_90": 0.19375, "token_acc": 0.9352909189777052, "epoch": 0.5101871790624483, "step": 1540}, {"loss": 0.17591552734375, "grad_norm": 0.47792771458625793, "learning_rate": 5.053646386207296e-06, "tIoU": 0.5278689861297607, "tRecall_50": 0.60625, "tRecall_75": 0.36875, "tRecall_90": 0.20625, "token_acc": 0.9347471451876019, "epoch": 0.5118436309425212, "step": 1545}, {"loss": 0.1619140625, "grad_norm": 0.4938163757324219, "learning_rate": 5.026823579100087e-06, "tIoU": 0.5453178882598877, "tRecall_50": 0.575, "tRecall_75": 0.4125, "tRecall_90": 0.275, "token_acc": 0.9415284199075333, "epoch": 0.513500082822594, "step": 1550}, {"loss": 0.16890869140625, "grad_norm": 0.7161099314689636, "learning_rate": 5e-06, "tIoU": 0.5169640064239502, "tRecall_50": 0.54375, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.9393939393939394, "epoch": 0.5151565347026669, "step": 1555}, {"loss": 0.16268310546875, "grad_norm": 0.5965278148651123, "learning_rate": 4.973176420899914e-06, "tIoU": 0.5604077339172363, "tRecall_50": 0.63125, "tRecall_75": 0.40625, "tRecall_90": 0.23125, "token_acc": 0.9370932754880694, "epoch": 0.5168129865827398, "step": 1560}, {"loss": 0.17548828125, "grad_norm": 1.2650108337402344, "learning_rate": 4.946353613792706e-06, "tIoU": 0.46562910079956055, "tRecall_50": 0.50625, "tRecall_75": 0.3625, "tRecall_90": 0.19375, "token_acc": 0.9333695947783519, "epoch": 0.5184694384628127, "step": 1565}, {"loss": 0.17598876953125, "grad_norm": 0.45799070596694946, "learning_rate": 4.919532350649033e-06, "tIoU": 0.5043049335479737, "tRecall_50": 0.55625, "tRecall_75": 0.375, "tRecall_90": 0.175, "token_acc": 0.9324434758921275, "epoch": 0.5201258903428856, "step": 1570}, {"loss": 0.17666015625, "grad_norm": 0.6668441891670227, "learning_rate": 4.8927134033951224e-06, "tIoU": 0.4904634475708008, "tRecall_50": 0.50625, "tRecall_75": 0.3875, "tRecall_90": 0.225, "token_acc": 0.9364130434782608, "epoch": 0.5217823422229584, "step": 1575}, {"loss": 0.155224609375, "grad_norm": 0.43300360441207886, "learning_rate": 4.865897543890541e-06, "tIoU": 0.5382334232330322, "tRecall_50": 0.5875, "tRecall_75": 0.375, "tRecall_90": 0.24375, "token_acc": 0.9438970990695129, "epoch": 0.5234387941030313, "step": 1580}, {"loss": 0.179248046875, "grad_norm": 0.6325023174285889, "learning_rate": 4.8390855439059955e-06, "tIoU": 0.46796865463256837, "tRecall_50": 0.51875, "tRecall_75": 0.2875, "tRecall_90": 0.15, "token_acc": 0.9322633297062024, "epoch": 0.5250952459831042, "step": 1585}, {"loss": 0.178515625, "grad_norm": 0.49443551898002625, "learning_rate": 4.812278175101112e-06, "tIoU": 0.5053191661834717, "tRecall_50": 0.50625, "tRecall_75": 0.38125, "tRecall_90": 0.25625, "token_acc": 0.9367157665030006, "epoch": 0.5267516978631771, "step": 1590}, {"loss": 0.17664794921875, "grad_norm": 0.5553008317947388, "learning_rate": 4.7854762090022274e-06, "tIoU": 0.49562931060791016, "tRecall_50": 0.55, "tRecall_75": 0.34375, "tRecall_90": 0.1875, "token_acc": 0.9352557127312296, "epoch": 0.52840814974325, "step": 1595}, {"loss": 0.16912841796875, "grad_norm": 0.825242817401886, "learning_rate": 4.758680416980193e-06, "tIoU": 0.4986118793487549, "tRecall_50": 0.55625, "tRecall_75": 0.325, "tRecall_90": 0.18125, "token_acc": 0.9382984506659419, "epoch": 0.5300646016233228, "step": 1600}, {"loss": 0.175537109375, "grad_norm": 0.5749291777610779, "learning_rate": 4.731891570228159e-06, "tIoU": 0.46056160926818845, "tRecall_50": 0.50625, "tRecall_75": 0.3625, "tRecall_90": 0.18125, "token_acc": 0.9333694474539544, "epoch": 0.5317210535033957, "step": 1605}, {"loss": 0.1766845703125, "grad_norm": 0.5613634586334229, "learning_rate": 4.7051104397393974e-06, "tIoU": 0.4753167152404785, "tRecall_50": 0.5, "tRecall_75": 0.325, "tRecall_90": 0.18125, "token_acc": 0.9323737099402498, "epoch": 0.5333775053834686, "step": 1610}, {"loss": 0.17596435546875, "grad_norm": 0.5255857110023499, "learning_rate": 4.678337796285093e-06, "tIoU": 0.5120919704437256, "tRecall_50": 0.55625, "tRecall_75": 0.3875, "tRecall_90": 0.23125, "token_acc": 0.9383692391600763, "epoch": 0.5350339572635415, "step": 1615}, {"loss": 0.16297607421875, "grad_norm": 0.7099191546440125, "learning_rate": 4.6515744103921786e-06, "tIoU": 0.5286399841308593, "tRecall_50": 0.58125, "tRecall_75": 0.3625, "tRecall_90": 0.25625, "token_acc": 0.9428024083196497, "epoch": 0.5366904091436144, "step": 1620}, {"loss": 0.16353759765625, "grad_norm": 0.5085911154747009, "learning_rate": 4.6248210523211465e-06, "tIoU": 0.5484652042388916, "tRecall_50": 0.60625, "tRecall_75": 0.36875, "tRecall_90": 0.2125, "token_acc": 0.9333151878062058, "epoch": 0.5383468610236872, "step": 1625}, {"loss": 0.16826171875, "grad_norm": 0.7192174792289734, "learning_rate": 4.5980784920438886e-06, "tIoU": 0.5181742668151855, "tRecall_50": 0.55, "tRecall_75": 0.40625, "tRecall_90": 0.21875, "token_acc": 0.9386536373507057, "epoch": 0.5400033129037601, "step": 1630}, {"loss": 0.1833984375, "grad_norm": 0.47222819924354553, "learning_rate": 4.571347499221528e-06, "tIoU": 0.4572866916656494, "tRecall_50": 0.4875, "tRecall_75": 0.28125, "tRecall_90": 0.175, "token_acc": 0.9301250679717238, "epoch": 0.541659764783833, "step": 1635}, {"loss": 0.16632080078125, "grad_norm": 0.6904683709144592, "learning_rate": 4.544628843182274e-06, "tIoU": 0.5224599838256836, "tRecall_50": 0.56875, "tRecall_75": 0.375, "tRecall_90": 0.2375, "token_acc": 0.9387755102040817, "epoch": 0.543316216663906, "step": 1640}, {"loss": 0.17559814453125, "grad_norm": 0.5094637274742126, "learning_rate": 4.51792329289928e-06, "tIoU": 0.5100062370300293, "tRecall_50": 0.50625, "tRecall_75": 0.36875, "tRecall_90": 0.2375, "token_acc": 0.9330966684871654, "epoch": 0.5449726685439787, "step": 1645}, {"loss": 0.168328857421875, "grad_norm": 0.38745856285095215, "learning_rate": 4.491231616968504e-06, "tIoU": 0.4967671871185303, "tRecall_50": 0.50625, "tRecall_75": 0.35, "tRecall_90": 0.2375, "token_acc": 0.937636761487965, "epoch": 0.5466291204240517, "step": 1650}, {"loss": 0.157330322265625, "grad_norm": 0.6616304516792297, "learning_rate": 4.464554583586605e-06, "tIoU": 0.5424611091613769, "tRecall_50": 0.5875, "tRecall_75": 0.3875, "tRecall_90": 0.275, "token_acc": 0.9404891304347827, "epoch": 0.5482855723041246, "step": 1655}, {"loss": 0.18017578125, "grad_norm": 3.821718215942383, "learning_rate": 4.437892960528815e-06, "tIoU": 0.5014461040496826, "tRecall_50": 0.5125, "tRecall_75": 0.3625, "tRecall_90": 0.25, "token_acc": 0.9336235038084875, "epoch": 0.5499420241841975, "step": 1660}, {"loss": 0.16463623046875, "grad_norm": 0.6461455225944519, "learning_rate": 4.411247515126857e-06, "tIoU": 0.4954252243041992, "tRecall_50": 0.54375, "tRecall_75": 0.3375, "tRecall_90": 0.2, "token_acc": 0.9375851730716817, "epoch": 0.5515984760642704, "step": 1665}, {"loss": 0.16251220703125, "grad_norm": 0.3840186297893524, "learning_rate": 4.384619014246848e-06, "tIoU": 0.5271695613861084, "tRecall_50": 0.55625, "tRecall_75": 0.3875, "tRecall_90": 0.25625, "token_acc": 0.9403882964178288, "epoch": 0.5532549279443432, "step": 1670}, {"loss": 0.15831298828125, "grad_norm": 0.6838501691818237, "learning_rate": 4.358008224267245e-06, "tIoU": 0.5436992168426513, "tRecall_50": 0.5875, "tRecall_75": 0.4375, "tRecall_90": 0.2375, "token_acc": 0.9403393541324576, "epoch": 0.5549113798244161, "step": 1675}, {"loss": 0.17099609375, "grad_norm": 0.48298609256744385, "learning_rate": 4.331415911056772e-06, "tIoU": 0.5099794864654541, "tRecall_50": 0.525, "tRecall_75": 0.4, "tRecall_90": 0.23125, "token_acc": 0.9344396082698585, "epoch": 0.556567831704489, "step": 1680}, {"loss": 0.18450927734375, "grad_norm": 0.6135856509208679, "learning_rate": 4.304842839952384e-06, "tIoU": 0.4841867446899414, "tRecall_50": 0.55, "tRecall_75": 0.3625, "tRecall_90": 0.16875, "token_acc": 0.9314658689148763, "epoch": 0.5582242835845619, "step": 1685}, {"loss": 0.16536865234375, "grad_norm": 0.5720863342285156, "learning_rate": 4.2782897757372414e-06, "tIoU": 0.5474737644195556, "tRecall_50": 0.6, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.9392778993435449, "epoch": 0.5598807354646348, "step": 1690}, {"loss": 0.17178955078125, "grad_norm": 0.5420379638671875, "learning_rate": 4.251757482618701e-06, "tIoU": 0.5092611312866211, "tRecall_50": 0.5625, "tRecall_75": 0.3625, "tRecall_90": 0.24375, "token_acc": 0.9342465753424658, "epoch": 0.5615371873447076, "step": 1695}, {"loss": 0.1845458984375, "grad_norm": 0.6336923837661743, "learning_rate": 4.22524672420632e-06, "tIoU": 0.4917005062103271, "tRecall_50": 0.55625, "tRecall_75": 0.3875, "tRecall_90": 0.2125, "token_acc": 0.9361702127659575, "epoch": 0.5631936392247805, "step": 1700}, {"loss": 0.18272705078125, "grad_norm": 0.9051311612129211, "learning_rate": 4.1987582634898724e-06, "tIoU": 0.5123229026794434, "tRecall_50": 0.58125, "tRecall_75": 0.36875, "tRecall_90": 0.20625, "token_acc": 0.9338820301783265, "epoch": 0.5648500911048534, "step": 1705}, {"loss": 0.160467529296875, "grad_norm": 0.7399612665176392, "learning_rate": 4.172292862817403e-06, "tIoU": 0.5589471340179444, "tRecall_50": 0.61875, "tRecall_75": 0.39375, "tRecall_90": 0.25, "token_acc": 0.9445506692160612, "epoch": 0.5665065429849263, "step": 1710}, {"loss": 0.18409423828125, "grad_norm": 0.41440925002098083, "learning_rate": 4.14585128387328e-06, "tIoU": 0.4675239086151123, "tRecall_50": 0.525, "tRecall_75": 0.35, "tRecall_90": 0.19375, "token_acc": 0.9282218597063622, "epoch": 0.5681629948649992, "step": 1715}, {"loss": 0.167669677734375, "grad_norm": 0.47274792194366455, "learning_rate": 4.119434287656265e-06, "tIoU": 0.5223277091979981, "tRecall_50": 0.5875, "tRecall_75": 0.35, "tRecall_90": 0.19375, "token_acc": 0.9380796508456083, "epoch": 0.569819446745072, "step": 1720}, {"loss": 0.162109375, "grad_norm": 0.49913594126701355, "learning_rate": 4.0930426344576255e-06, "tIoU": 0.45806708335876467, "tRecall_50": 0.50625, "tRecall_75": 0.325, "tRecall_90": 0.21875, "token_acc": 0.9379929290182214, "epoch": 0.5714758986251449, "step": 1725}, {"loss": 0.18353271484375, "grad_norm": 1.1948215961456299, "learning_rate": 4.0666770838392465e-06, "tIoU": 0.5288844585418702, "tRecall_50": 0.5625, "tRecall_75": 0.38125, "tRecall_90": 0.25, "token_acc": 0.9384363933533098, "epoch": 0.5731323505052178, "step": 1730}, {"loss": 0.1573974609375, "grad_norm": 0.4646267890930176, "learning_rate": 4.040338394611772e-06, "tIoU": 0.5586710453033448, "tRecall_50": 0.6125, "tRecall_75": 0.4125, "tRecall_90": 0.275, "token_acc": 0.941079295154185, "epoch": 0.5747888023852907, "step": 1735}, {"loss": 0.16785888671875, "grad_norm": 0.5275194048881531, "learning_rate": 4.014027324812761e-06, "tIoU": 0.5046965122222901, "tRecall_50": 0.5625, "tRecall_75": 0.38125, "tRecall_90": 0.20625, "token_acc": 0.9382682327232996, "epoch": 0.5764452542653636, "step": 1740}, {"loss": 0.1590087890625, "grad_norm": 0.5200470685958862, "learning_rate": 3.987744631684877e-06, "tIoU": 0.5480869293212891, "tRecall_50": 0.63125, "tRecall_75": 0.38125, "tRecall_90": 0.25, "token_acc": 0.9403919433859553, "epoch": 0.5781017061454364, "step": 1745}, {"loss": 0.174365234375, "grad_norm": 0.5766518712043762, "learning_rate": 3.961491071654095e-06, "tIoU": 0.508606243133545, "tRecall_50": 0.53125, "tRecall_75": 0.3375, "tRecall_90": 0.2375, "token_acc": 0.9362224039247752, "epoch": 0.5797581580255093, "step": 1750}, {"loss": 0.15159912109375, "grad_norm": 0.5985773801803589, "learning_rate": 3.9352674003079225e-06, "tIoU": 0.572104549407959, "tRecall_50": 0.65, "tRecall_75": 0.40625, "tRecall_90": 0.2625, "token_acc": 0.9412889131622064, "epoch": 0.5814146099055822, "step": 1755}, {"loss": 0.16317138671875, "grad_norm": 0.9872517585754395, "learning_rate": 3.909074372373662e-06, "tIoU": 0.5388112545013428, "tRecall_50": 0.58125, "tRecall_75": 0.425, "tRecall_90": 0.29375, "token_acc": 0.94379581862612, "epoch": 0.5830710617856552, "step": 1760}, {"loss": 0.14564208984375, "grad_norm": 0.5140849947929382, "learning_rate": 3.882912741696688e-06, "tIoU": 0.5485616683959961, "tRecall_50": 0.58125, "tRecall_75": 0.44375, "tRecall_90": 0.28125, "token_acc": 0.9443079443079443, "epoch": 0.5847275136657281, "step": 1765}, {"loss": 0.17509765625, "grad_norm": 0.6263390779495239, "learning_rate": 3.85678326121875e-06, "tIoU": 0.5056493282318115, "tRecall_50": 0.54375, "tRecall_75": 0.4, "tRecall_90": 0.21875, "token_acc": 0.9353518821603928, "epoch": 0.5863839655458009, "step": 1770}, {"loss": 0.17532958984375, "grad_norm": 0.6849157214164734, "learning_rate": 3.830686682956299e-06, "tIoU": 0.5150390148162842, "tRecall_50": 0.5875, "tRecall_75": 0.36875, "tRecall_90": 0.2125, "token_acc": 0.9324952172724789, "epoch": 0.5880404174258738, "step": 1775}, {"loss": 0.14559326171875, "grad_norm": 0.8499537110328674, "learning_rate": 3.804623757978849e-06, "tIoU": 0.5719325542449951, "tRecall_50": 0.625, "tRecall_75": 0.425, "tRecall_90": 0.325, "token_acc": 0.9457406402604449, "epoch": 0.5896968693059467, "step": 1780}, {"loss": 0.1645263671875, "grad_norm": 0.5121057033538818, "learning_rate": 3.778595236387367e-06, "tIoU": 0.5143845081329346, "tRecall_50": 0.5625, "tRecall_75": 0.40625, "tRecall_90": 0.23125, "token_acc": 0.9362224039247752, "epoch": 0.5913533211860196, "step": 1785}, {"loss": 0.17528076171875, "grad_norm": 0.7534790635108948, "learning_rate": 3.752601867292664e-06, "tIoU": 0.4931938648223877, "tRecall_50": 0.5375, "tRecall_75": 0.31875, "tRecall_90": 0.16875, "token_acc": 0.9328256731030732, "epoch": 0.5930097730660925, "step": 1790}, {"loss": 0.16246337890625, "grad_norm": 0.5419090986251831, "learning_rate": 3.726644398793857e-06, "tIoU": 0.5456307888031006, "tRecall_50": 0.6375, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.9386536373507057, "epoch": 0.5946662249461653, "step": 1795}, {"loss": 0.18109130859375, "grad_norm": 0.5578802824020386, "learning_rate": 3.7007235779568303e-06, "tIoU": 0.4708672046661377, "tRecall_50": 0.51875, "tRecall_75": 0.29375, "tRecall_90": 0.14375, "token_acc": 0.9316146540027137, "epoch": 0.5963226768262382, "step": 1800}, {"loss": 0.169842529296875, "grad_norm": 1.2091337442398071, "learning_rate": 3.6748401507927327e-06, "tIoU": 0.49648056030273435, "tRecall_50": 0.525, "tRecall_75": 0.35625, "tRecall_90": 0.2125, "token_acc": 0.9373644251626898, "epoch": 0.5979791287063111, "step": 1805}, {"loss": 0.16412353515625, "grad_norm": 0.6194023489952087, "learning_rate": 3.648994862236508e-06, "tIoU": 0.5116681098937989, "tRecall_50": 0.58125, "tRecall_75": 0.35625, "tRecall_90": 0.175, "token_acc": 0.9343724364232978, "epoch": 0.599635580586384, "step": 1810}, {"loss": 0.1682373046875, "grad_norm": 0.5854647159576416, "learning_rate": 3.6231884561254582e-06, "tIoU": 0.516246223449707, "tRecall_50": 0.58125, "tRecall_75": 0.35, "tRecall_90": 0.21875, "token_acc": 0.9384531590413944, "epoch": 0.6012920324664568, "step": 1815}, {"loss": 0.1697265625, "grad_norm": 0.7405495643615723, "learning_rate": 3.5974216751778325e-06, "tIoU": 0.5363904476165772, "tRecall_50": 0.58125, "tRecall_75": 0.4125, "tRecall_90": 0.21875, "token_acc": 0.9363784665579119, "epoch": 0.6029484843465297, "step": 1820}, {"loss": 0.1662109375, "grad_norm": 0.45874157547950745, "learning_rate": 3.5716952609714517e-06, "tIoU": 0.5525872707366943, "tRecall_50": 0.61875, "tRecall_75": 0.38125, "tRecall_90": 0.2125, "token_acc": 0.9399188092016239, "epoch": 0.6046049362266026, "step": 1825}, {"loss": 0.164111328125, "grad_norm": 0.6839113235473633, "learning_rate": 3.546009953922368e-06, "tIoU": 0.49982576370239257, "tRecall_50": 0.51875, "tRecall_75": 0.325, "tRecall_90": 0.20625, "token_acc": 0.9340419732897247, "epoch": 0.6062613881066755, "step": 1830}, {"loss": 0.1625, "grad_norm": 0.6627931594848633, "learning_rate": 3.520366493263554e-06, "tIoU": 0.5280773162841796, "tRecall_50": 0.575, "tRecall_75": 0.375, "tRecall_90": 0.2375, "token_acc": 0.9365641165260006, "epoch": 0.6079178399867484, "step": 1835}, {"loss": 0.1634521484375, "grad_norm": 0.8185617327690125, "learning_rate": 3.4947656170236255e-06, "tIoU": 0.4996767520904541, "tRecall_50": 0.5375, "tRecall_75": 0.34375, "tRecall_90": 0.2, "token_acc": 0.9390710382513662, "epoch": 0.6095742918668212, "step": 1840}, {"loss": 0.1591796875, "grad_norm": 0.4964151382446289, "learning_rate": 3.4692080620056044e-06, "tIoU": 0.5419831275939941, "tRecall_50": 0.58125, "tRecall_75": 0.40625, "tRecall_90": 0.28125, "token_acc": 0.941464742717125, "epoch": 0.6112307437468941, "step": 1845}, {"loss": 0.18828125, "grad_norm": 0.7755962610244751, "learning_rate": 3.4436945637657106e-06, "tIoU": 0.4821001052856445, "tRecall_50": 0.5125, "tRecall_75": 0.36875, "tRecall_90": 0.175, "token_acc": 0.9328785811732606, "epoch": 0.612887195626967, "step": 1850}, {"loss": 0.1717041015625, "grad_norm": 0.42528417706489563, "learning_rate": 3.4182258565921933e-06, "tIoU": 0.5007997512817383, "tRecall_50": 0.5125, "tRecall_75": 0.33125, "tRecall_90": 0.19375, "token_acc": 0.9337658826709921, "epoch": 0.6145436475070399, "step": 1855}, {"loss": 0.15889892578125, "grad_norm": 0.549651563167572, "learning_rate": 3.3928026734841935e-06, "tIoU": 0.5112274646759033, "tRecall_50": 0.54375, "tRecall_75": 0.375, "tRecall_90": 0.24375, "token_acc": 0.9406641262928688, "epoch": 0.6162000993871128, "step": 1860}, {"loss": 0.167706298828125, "grad_norm": 0.5841197967529297, "learning_rate": 3.367425746130658e-06, "tIoU": 0.510018253326416, "tRecall_50": 0.55, "tRecall_75": 0.3375, "tRecall_90": 0.2375, "token_acc": 0.9364949577541565, "epoch": 0.6178565512671856, "step": 1865}, {"loss": 0.16390380859375, "grad_norm": 0.8597052097320557, "learning_rate": 3.342095804889275e-06, "tIoU": 0.5023562431335449, "tRecall_50": 0.5625, "tRecall_75": 0.3625, "tRecall_90": 0.2, "token_acc": 0.9369713506139155, "epoch": 0.6195130031472585, "step": 1870}, {"loss": 0.17318115234375, "grad_norm": 0.4235498011112213, "learning_rate": 3.3168135787654493e-06, "tIoU": 0.4791114807128906, "tRecall_50": 0.5375, "tRecall_75": 0.3125, "tRecall_90": 0.15625, "token_acc": 0.934092758340114, "epoch": 0.6211694550273315, "step": 1875}, {"loss": 0.1798095703125, "grad_norm": 0.5873208045959473, "learning_rate": 3.291579795391331e-06, "tIoU": 0.5017872333526612, "tRecall_50": 0.55, "tRecall_75": 0.375, "tRecall_90": 0.2375, "token_acc": 0.9341496598639456, "epoch": 0.6228259069074044, "step": 1880}, {"loss": 0.16212158203125, "grad_norm": 0.27464815974235535, "learning_rate": 3.2663951810048683e-06, "tIoU": 0.5321149826049805, "tRecall_50": 0.58125, "tRecall_75": 0.3875, "tRecall_90": 0.25625, "token_acc": 0.9363162228932199, "epoch": 0.6244823587874773, "step": 1885}, {"loss": 0.162310791015625, "grad_norm": 0.5738742351531982, "learning_rate": 3.2412604604289073e-06, "tIoU": 0.5201348781585693, "tRecall_50": 0.56875, "tRecall_75": 0.3875, "tRecall_90": 0.23125, "token_acc": 0.9401360544217687, "epoch": 0.6261388106675501, "step": 1890}, {"loss": 0.181396484375, "grad_norm": 0.44134360551834106, "learning_rate": 3.216176357050327e-06, "tIoU": 0.4499537467956543, "tRecall_50": 0.4625, "tRecall_75": 0.325, "tRecall_90": 0.2, "token_acc": 0.9313432835820895, "epoch": 0.627795262547623, "step": 1895}, {"loss": 0.1660888671875, "grad_norm": 0.49762654304504395, "learning_rate": 3.1911435927992298e-06, "tIoU": 0.5242837905883789, "tRecall_50": 0.54375, "tRecall_75": 0.375, "tRecall_90": 0.25625, "token_acc": 0.9374828626268166, "epoch": 0.6294517144276959, "step": 1900}, {"loss": 0.156964111328125, "grad_norm": 0.6275957822799683, "learning_rate": 3.1661628881281564e-06, "tIoU": 0.562921667098999, "tRecall_50": 0.59375, "tRecall_75": 0.46875, "tRecall_90": 0.3, "token_acc": 0.9451119606772256, "epoch": 0.6311081663077688, "step": 1905}, {"loss": 0.17091064453125, "grad_norm": 0.5966311693191528, "learning_rate": 3.1412349619913514e-06, "tIoU": 0.5571235179901123, "tRecall_50": 0.6, "tRecall_75": 0.45, "tRecall_90": 0.2625, "token_acc": 0.9354223433242507, "epoch": 0.6327646181878417, "step": 1910}, {"loss": 0.16427001953125, "grad_norm": 0.4650852084159851, "learning_rate": 3.116360531824074e-06, "tIoU": 0.5082977294921875, "tRecall_50": 0.55625, "tRecall_75": 0.34375, "tRecall_90": 0.2125, "token_acc": 0.9358449358449359, "epoch": 0.6344210700679145, "step": 1915}, {"loss": 0.16798095703125, "grad_norm": 0.534633457660675, "learning_rate": 3.091540313521952e-06, "tIoU": 0.493639612197876, "tRecall_50": 0.525, "tRecall_75": 0.375, "tRecall_90": 0.26875, "token_acc": 0.9388145315487572, "epoch": 0.6360775219479874, "step": 1920}, {"loss": 0.1730224609375, "grad_norm": 0.3949797749519348, "learning_rate": 3.0667750214203697e-06, "tIoU": 0.5319153308868408, "tRecall_50": 0.58125, "tRecall_75": 0.40625, "tRecall_90": 0.2625, "token_acc": 0.9356328082563824, "epoch": 0.6377339738280603, "step": 1925}, {"loss": 0.18037109375, "grad_norm": 0.7183004021644592, "learning_rate": 3.0420653682739154e-06, "tIoU": 0.4745781421661377, "tRecall_50": 0.51875, "tRecall_75": 0.3, "tRecall_90": 0.16875, "token_acc": 0.9310910582444627, "epoch": 0.6393904257081332, "step": 1930}, {"loss": 0.16229248046875, "grad_norm": 0.4463639557361603, "learning_rate": 3.0174120652358683e-06, "tIoU": 0.52823486328125, "tRecall_50": 0.55625, "tRecall_75": 0.36875, "tRecall_90": 0.24375, "token_acc": 0.9381302807304442, "epoch": 0.6410468775882061, "step": 1935}, {"loss": 0.171685791015625, "grad_norm": 0.812966525554657, "learning_rate": 2.992815821837732e-06, "tIoU": 0.504635238647461, "tRecall_50": 0.5375, "tRecall_75": 0.375, "tRecall_90": 0.25625, "token_acc": 0.9380771319934817, "epoch": 0.6427033294682789, "step": 1940}, {"loss": 0.163983154296875, "grad_norm": 0.5824111104011536, "learning_rate": 2.9682773459688087e-06, "tIoU": 0.5575588703155517, "tRecall_50": 0.60625, "tRecall_75": 0.40625, "tRecall_90": 0.25625, "token_acc": 0.9411442649876813, "epoch": 0.6443597813483518, "step": 1945}, {"loss": 0.15794677734375, "grad_norm": 0.754898190498352, "learning_rate": 2.943797343855829e-06, "tIoU": 0.5148393154144287, "tRecall_50": 0.55, "tRecall_75": 0.425, "tRecall_90": 0.23125, "token_acc": 0.9404567699836868, "epoch": 0.6460162332284247, "step": 1950}, {"loss": 0.16236572265625, "grad_norm": 0.549222469329834, "learning_rate": 2.919376520042628e-06, "tIoU": 0.5246635913848877, "tRecall_50": 0.575, "tRecall_75": 0.4, "tRecall_90": 0.20625, "token_acc": 0.9405860010851872, "epoch": 0.6476726851084976, "step": 1955}, {"loss": 0.15792236328125, "grad_norm": 0.5977454781532288, "learning_rate": 2.895015577369866e-06, "tIoU": 0.5148863315582275, "tRecall_50": 0.56875, "tRecall_75": 0.3875, "tRecall_90": 0.23125, "token_acc": 0.9405214557305812, "epoch": 0.6493291369885705, "step": 1960}, {"loss": 0.15865478515625, "grad_norm": 0.40176132321357727, "learning_rate": 2.8707152169548014e-06, "tIoU": 0.5105422973632813, "tRecall_50": 0.5375, "tRecall_75": 0.35625, "tRecall_90": 0.2625, "token_acc": 0.9393527332064183, "epoch": 0.6509855888686433, "step": 1965}, {"loss": 0.15938720703125, "grad_norm": 0.6446389555931091, "learning_rate": 2.846476138171111e-06, "tIoU": 0.5290540218353271, "tRecall_50": 0.5625, "tRecall_75": 0.36875, "tRecall_90": 0.275, "token_acc": 0.9428882240957303, "epoch": 0.6526420407487162, "step": 1970}, {"loss": 0.169287109375, "grad_norm": 0.41462159156799316, "learning_rate": 2.822299038628762e-06, "tIoU": 0.5018656253814697, "tRecall_50": 0.4875, "tRecall_75": 0.35625, "tRecall_90": 0.25625, "token_acc": 0.9404081632653061, "epoch": 0.6542984926287891, "step": 1975}, {"loss": 0.16522216796875, "grad_norm": 0.9564310908317566, "learning_rate": 2.7981846141539402e-06, "tIoU": 0.5413733005523682, "tRecall_50": 0.6, "tRecall_75": 0.39375, "tRecall_90": 0.2625, "token_acc": 0.9384363933533098, "epoch": 0.655954944508862, "step": 1980}, {"loss": 0.16549072265625, "grad_norm": 0.600674569606781, "learning_rate": 2.774133558769014e-06, "tIoU": 0.5427555561065673, "tRecall_50": 0.6125, "tRecall_75": 0.3875, "tRecall_90": 0.275, "token_acc": 0.9395336225596529, "epoch": 0.657611396388935, "step": 1985}, {"loss": 0.156494140625, "grad_norm": 0.6821838021278381, "learning_rate": 2.7501465646725646e-06, "tIoU": 0.5400360584259033, "tRecall_50": 0.5625, "tRecall_75": 0.4125, "tRecall_90": 0.28125, "token_acc": 0.9448689956331878, "epoch": 0.6592678482690078, "step": 1990}, {"loss": 0.15616455078125, "grad_norm": 0.5166508555412292, "learning_rate": 2.7262243222194728e-06, "tIoU": 0.535555648803711, "tRecall_50": 0.60625, "tRecall_75": 0.36875, "tRecall_90": 0.21875, "token_acc": 0.9406130268199234, "epoch": 0.6609243001490807, "step": 1995}, {"loss": 0.17166748046875, "grad_norm": 0.7035804390907288, "learning_rate": 2.70236751990103e-06, "tIoU": 0.5096180438995361, "tRecall_50": 0.54375, "tRecall_75": 0.3625, "tRecall_90": 0.20625, "token_acc": 0.9358172423171064, "epoch": 0.6625807520291536, "step": 2000}, {"loss": 0.161578369140625, "grad_norm": 0.5203202962875366, "learning_rate": 2.6785768443251437e-06, "tIoU": 0.5287741184234619, "tRecall_50": 0.5625, "tRecall_75": 0.35625, "tRecall_90": 0.23125, "token_acc": 0.9400934322616104, "epoch": 0.6642372039092265, "step": 2005}, {"loss": 0.168927001953125, "grad_norm": 0.8393142819404602, "learning_rate": 2.6548529801965683e-06, "tIoU": 0.5189241886138916, "tRecall_50": 0.55625, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.9400328587075575, "epoch": 0.6658936557892993, "step": 2010}, {"loss": 0.16002197265625, "grad_norm": 0.4761408865451813, "learning_rate": 2.6311966102971964e-06, "tIoU": 0.49315404891967773, "tRecall_50": 0.55, "tRecall_75": 0.35625, "tRecall_90": 0.2625, "token_acc": 0.9369197160021846, "epoch": 0.6675501076693722, "step": 2015}, {"loss": 0.163818359375, "grad_norm": 0.638813316822052, "learning_rate": 2.60760841546641e-06, "tIoU": 0.5471360683441162, "tRecall_50": 0.5875, "tRecall_75": 0.4, "tRecall_90": 0.26875, "token_acc": 0.9425006780580417, "epoch": 0.6692065595494451, "step": 2020}, {"loss": 0.17288818359375, "grad_norm": 0.47899043560028076, "learning_rate": 2.5840890745814927e-06, "tIoU": 0.49503450393676757, "tRecall_50": 0.53125, "tRecall_75": 0.33125, "tRecall_90": 0.2625, "token_acc": 0.9383019383019383, "epoch": 0.670863011429518, "step": 2025}, {"loss": 0.16365966796875, "grad_norm": 1.0909929275512695, "learning_rate": 2.5606392645380783e-06, "tIoU": 0.5265616416931153, "tRecall_50": 0.56875, "tRecall_75": 0.4, "tRecall_90": 0.25, "token_acc": 0.9414980863860033, "epoch": 0.6725194633095909, "step": 2030}, {"loss": 0.15804443359375, "grad_norm": 0.57685786485672, "learning_rate": 2.537259660230679e-06, "tIoU": 0.5010174751281739, "tRecall_50": 0.53125, "tRecall_75": 0.4, "tRecall_90": 0.23125, "token_acc": 0.9364776444929117, "epoch": 0.6741759151896637, "step": 2035}, {"loss": 0.1663818359375, "grad_norm": 0.8344061970710754, "learning_rate": 2.513950934533259e-06, "tIoU": 0.5199985027313232, "tRecall_50": 0.50625, "tRecall_75": 0.40625, "tRecall_90": 0.28125, "token_acc": 0.9410801963993454, "epoch": 0.6758323670697366, "step": 2040}, {"loss": 0.1641357421875, "grad_norm": 0.7436922192573547, "learning_rate": 2.4907137582798676e-06, "tIoU": 0.4792342185974121, "tRecall_50": 0.5125, "tRecall_75": 0.3375, "tRecall_90": 0.225, "token_acc": 0.9375851730716817, "epoch": 0.6774888189498095, "step": 2045}, {"loss": 0.1564208984375, "grad_norm": 0.5653188824653625, "learning_rate": 2.4675488002453385e-06, "tIoU": 0.5584266185760498, "tRecall_50": 0.6, "tRecall_75": 0.4625, "tRecall_90": 0.2875, "token_acc": 0.9416098226466576, "epoch": 0.6791452708298824, "step": 2050}, {"loss": 0.17310791015625, "grad_norm": 0.5735906362533569, "learning_rate": 2.4444567271260316e-06, "tIoU": 0.49056329727172854, "tRecall_50": 0.54375, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.9352654387865655, "epoch": 0.6808017227099553, "step": 2055}, {"loss": 0.1667236328125, "grad_norm": 0.734527051448822, "learning_rate": 2.4214382035206525e-06, "tIoU": 0.5110460758209229, "tRecall_50": 0.56875, "tRecall_75": 0.35, "tRecall_90": 0.2, "token_acc": 0.9353695118625579, "epoch": 0.6824581745900281, "step": 2060}, {"loss": 0.15985107421875, "grad_norm": 1.2782374620437622, "learning_rate": 2.398493891911127e-06, "tIoU": 0.564049768447876, "tRecall_50": 0.60625, "tRecall_75": 0.44375, "tRecall_90": 0.2625, "token_acc": 0.9417555373256767, "epoch": 0.684114626470101, "step": 2065}, {"loss": 0.164990234375, "grad_norm": 0.5806080102920532, "learning_rate": 2.3756244526435303e-06, "tIoU": 0.49380931854248045, "tRecall_50": 0.54375, "tRecall_75": 0.3625, "tRecall_90": 0.24375, "token_acc": 0.9363091997822537, "epoch": 0.6857710783501739, "step": 2070}, {"loss": 0.166650390625, "grad_norm": 0.599310040473938, "learning_rate": 2.3528305439090744e-06, "tIoU": 0.5209931373596192, "tRecall_50": 0.575, "tRecall_75": 0.38125, "tRecall_90": 0.2375, "token_acc": 0.9398047722342733, "epoch": 0.6874275302302468, "step": 2075}, {"loss": 0.16470947265625, "grad_norm": 0.6360598802566528, "learning_rate": 2.3301128217251866e-06, "tIoU": 0.5022995948791504, "tRecall_50": 0.50625, "tRecall_75": 0.36875, "tRecall_90": 0.26875, "token_acc": 0.9365773646801531, "epoch": 0.6890839821103197, "step": 2080}, {"loss": 0.166156005859375, "grad_norm": 0.6225826144218445, "learning_rate": 2.3074719399166085e-06, "tIoU": 0.532801342010498, "tRecall_50": 0.55625, "tRecall_75": 0.4, "tRecall_90": 0.26875, "token_acc": 0.9410962639760022, "epoch": 0.6907404339903925, "step": 2085}, {"loss": 0.1636474609375, "grad_norm": 0.543380081653595, "learning_rate": 2.284908550096586e-06, "tIoU": 0.49683642387390137, "tRecall_50": 0.5375, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.9385489983757445, "epoch": 0.6923968858704654, "step": 2090}, {"loss": 0.168560791015625, "grad_norm": 0.6770683526992798, "learning_rate": 2.2624233016481224e-06, "tIoU": 0.4816957950592041, "tRecall_50": 0.5125, "tRecall_75": 0.35625, "tRecall_90": 0.2375, "token_acc": 0.9354223433242507, "epoch": 0.6940533377505383, "step": 2095}, {"loss": 0.1667724609375, "grad_norm": 0.49540141224861145, "learning_rate": 2.240016841705274e-06, "tIoU": 0.5249676704406738, "tRecall_50": 0.55625, "tRecall_75": 0.4, "tRecall_90": 0.23125, "token_acc": 0.940424374319913, "epoch": 0.6957097896306113, "step": 2100}, {"loss": 0.1658203125, "grad_norm": 0.5058975219726562, "learning_rate": 2.217689815134539e-06, "tIoU": 0.5202480792999268, "tRecall_50": 0.54375, "tRecall_75": 0.38125, "tRecall_90": 0.25, "token_acc": 0.9400054540496319, "epoch": 0.6973662415106842, "step": 2105}, {"loss": 0.164404296875, "grad_norm": 0.3797245919704437, "learning_rate": 2.1954428645162914e-06, "tIoU": 0.5144438743591309, "tRecall_50": 0.54375, "tRecall_75": 0.4125, "tRecall_90": 0.2625, "token_acc": 0.9379084967320261, "epoch": 0.699022693390757, "step": 2110}, {"loss": 0.1622802734375, "grad_norm": 0.6059390306472778, "learning_rate": 2.173276630126287e-06, "tIoU": 0.5094360828399658, "tRecall_50": 0.5625, "tRecall_75": 0.3625, "tRecall_90": 0.225, "token_acc": 0.9365986394557824, "epoch": 0.7006791452708299, "step": 2115}, {"loss": 0.16075439453125, "grad_norm": 0.7819738984107971, "learning_rate": 2.151191749917239e-06, "tIoU": 0.5810178279876709, "tRecall_50": 0.6125, "tRecall_75": 0.43125, "tRecall_90": 0.3125, "token_acc": 0.9424754634678298, "epoch": 0.7023355971509028, "step": 2120}, {"loss": 0.16397705078125, "grad_norm": 0.5492057800292969, "learning_rate": 2.129188859500459e-06, "tIoU": 0.5358535289764405, "tRecall_50": 0.575, "tRecall_75": 0.39375, "tRecall_90": 0.225, "token_acc": 0.9394104803493449, "epoch": 0.7039920490309757, "step": 2125}, {"loss": 0.163037109375, "grad_norm": 0.5017256736755371, "learning_rate": 2.107268592127557e-06, "tIoU": 0.5214012145996094, "tRecall_50": 0.60625, "tRecall_75": 0.31875, "tRecall_90": 0.2, "token_acc": 0.9364949577541565, "epoch": 0.7056485009110486, "step": 2130}, {"loss": 0.17164306640625, "grad_norm": 0.6092507839202881, "learning_rate": 2.085431578672219e-06, "tIoU": 0.5037929058074951, "tRecall_50": 0.58125, "tRecall_75": 0.33125, "tRecall_90": 0.2, "token_acc": 0.9367502726281353, "epoch": 0.7073049527911214, "step": 2135}, {"loss": 0.16177978515625, "grad_norm": 0.5345575213432312, "learning_rate": 2.0636784476120598e-06, "tIoU": 0.5068227291107178, "tRecall_50": 0.55, "tRecall_75": 0.39375, "tRecall_90": 0.20625, "token_acc": 0.940424374319913, "epoch": 0.7089614046711943, "step": 2140}, {"loss": 0.161236572265625, "grad_norm": 0.8379113674163818, "learning_rate": 2.042009825010513e-06, "tIoU": 0.5136396884918213, "tRecall_50": 0.51875, "tRecall_75": 0.35625, "tRecall_90": 0.2375, "token_acc": 0.9387087986924544, "epoch": 0.7106178565512672, "step": 2145}, {"loss": 0.1553955078125, "grad_norm": 0.5098614692687988, "learning_rate": 2.0204263344988395e-06, "tIoU": 0.573584508895874, "tRecall_50": 0.6375, "tRecall_75": 0.4125, "tRecall_90": 0.3, "token_acc": 0.941320293398533, "epoch": 0.7122743084313401, "step": 2150}, {"loss": 0.16419677734375, "grad_norm": 0.7528647184371948, "learning_rate": 1.9989285972581595e-06, "tIoU": 0.5330680847167969, "tRecall_50": 0.5625, "tRecall_75": 0.4, "tRecall_90": 0.21875, "token_acc": 0.9387254901960784, "epoch": 0.713930760311413, "step": 2155}, {"loss": 0.1602294921875, "grad_norm": 0.47473061084747314, "learning_rate": 1.9775172320015805e-06, "tIoU": 0.5205780506134033, "tRecall_50": 0.58125, "tRecall_75": 0.3875, "tRecall_90": 0.2, "token_acc": 0.9394843962008141, "epoch": 0.7155872121914858, "step": 2160}, {"loss": 0.17667236328125, "grad_norm": 0.9584677219390869, "learning_rate": 1.956192854956397e-06, "tIoU": 0.47919406890869143, "tRecall_50": 0.50625, "tRecall_75": 0.38125, "tRecall_90": 0.21875, "token_acc": 0.9352750809061489, "epoch": 0.7172436640715587, "step": 2165}, {"loss": 0.145880126953125, "grad_norm": 0.5655807852745056, "learning_rate": 1.9349560798463423e-06, "tIoU": 0.5800382137298584, "tRecall_50": 0.6125, "tRecall_75": 0.45, "tRecall_90": 0.28125, "token_acc": 0.9449164154562893, "epoch": 0.7189001159516316, "step": 2170}, {"loss": 0.15904541015625, "grad_norm": 0.4855189323425293, "learning_rate": 1.9138075178739375e-06, "tIoU": 0.5785966873168945, "tRecall_50": 0.6375, "tRecall_75": 0.4625, "tRecall_90": 0.31875, "token_acc": 0.9443683200876952, "epoch": 0.7205565678317045, "step": 2175}, {"loss": 0.16480712890625, "grad_norm": 0.42149361968040466, "learning_rate": 1.8927477777028934e-06, "tIoU": 0.5444125652313232, "tRecall_50": 0.56875, "tRecall_75": 0.40625, "tRecall_90": 0.2625, "token_acc": 0.9375678610206297, "epoch": 0.7222130197117774, "step": 2180}, {"loss": 0.151251220703125, "grad_norm": 0.8240121006965637, "learning_rate": 1.8717774654405962e-06, "tIoU": 0.5565950393676757, "tRecall_50": 0.6, "tRecall_75": 0.41875, "tRecall_90": 0.2875, "token_acc": 0.9450013546464373, "epoch": 0.7238694715918502, "step": 2185}, {"loss": 0.1607177734375, "grad_norm": 0.41935020685195923, "learning_rate": 1.8508971846206603e-06, "tIoU": 0.5410372257232666, "tRecall_50": 0.575, "tRecall_75": 0.40625, "tRecall_90": 0.30625, "token_acc": 0.9429812652728754, "epoch": 0.7255259234719231, "step": 2190}, {"loss": 0.1631591796875, "grad_norm": 0.6102359294891357, "learning_rate": 1.8301075361855674e-06, "tIoU": 0.49234714508056643, "tRecall_50": 0.525, "tRecall_75": 0.3375, "tRecall_90": 0.25, "token_acc": 0.9369517543859649, "epoch": 0.727182375351996, "step": 2195}, {"loss": 0.1391357421875, "grad_norm": 0.599194347858429, "learning_rate": 1.8094091184693569e-06, "tIoU": 0.5903083324432373, "tRecall_50": 0.61875, "tRecall_75": 0.49375, "tRecall_90": 0.30625, "token_acc": 0.9462336244541485, "epoch": 0.7288388272320689, "step": 2200}, {"loss": 0.1526611328125, "grad_norm": 0.6685839295387268, "learning_rate": 1.7888025271804156e-06, "tIoU": 0.5808693408966065, "tRecall_50": 0.60625, "tRecall_75": 0.4375, "tRecall_90": 0.3375, "token_acc": 0.944063613929257, "epoch": 0.7304952791121417, "step": 2205}, {"loss": 0.16241455078125, "grad_norm": 0.5242391228675842, "learning_rate": 1.7682883553843356e-06, "tIoU": 0.5209659576416016, "tRecall_50": 0.5625, "tRecall_75": 0.39375, "tRecall_90": 0.24375, "token_acc": 0.9373126192423004, "epoch": 0.7321517309922146, "step": 2210}, {"loss": 0.16380615234375, "grad_norm": 0.6233892440795898, "learning_rate": 1.7478671934868302e-06, "tIoU": 0.5173926830291748, "tRecall_50": 0.55625, "tRecall_75": 0.3875, "tRecall_90": 0.225, "token_acc": 0.9400492745688476, "epoch": 0.7338081828722876, "step": 2215}, {"loss": 0.1623779296875, "grad_norm": 0.8942482471466064, "learning_rate": 1.7275396292167623e-06, "tIoU": 0.5592443943023682, "tRecall_50": 0.6, "tRecall_75": 0.4, "tRecall_90": 0.28125, "token_acc": 0.943504366812227, "epoch": 0.7354646347523605, "step": 2220}, {"loss": 0.174151611328125, "grad_norm": 0.7321513891220093, "learning_rate": 1.7073062476092145e-06, "tIoU": 0.4566510200500488, "tRecall_50": 0.4875, "tRecall_75": 0.33125, "tRecall_90": 0.14375, "token_acc": 0.9328804347826087, "epoch": 0.7371210866324334, "step": 2225}, {"loss": 0.1498046875, "grad_norm": 0.6549537181854248, "learning_rate": 1.6871676309886548e-06, "tIoU": 0.5559914588928223, "tRecall_50": 0.59375, "tRecall_75": 0.41875, "tRecall_90": 0.25, "token_acc": 0.9419496166484118, "epoch": 0.7387775385125062, "step": 2230}, {"loss": 0.16259765625, "grad_norm": 0.6222982406616211, "learning_rate": 1.667124358952184e-06, "tIoU": 0.5121729850769043, "tRecall_50": 0.50625, "tRecall_75": 0.35625, "tRecall_90": 0.2375, "token_acc": 0.9351851851851852, "epoch": 0.7404339903925791, "step": 2235}, {"loss": 0.16424560546875, "grad_norm": 0.6524860858917236, "learning_rate": 1.6471770083528444e-06, "tIoU": 0.5048126697540283, "tRecall_50": 0.54375, "tRecall_75": 0.35625, "tRecall_90": 0.20625, "token_acc": 0.9371273712737127, "epoch": 0.742090442272652, "step": 2240}, {"loss": 0.165899658203125, "grad_norm": 0.6570249199867249, "learning_rate": 1.6273261532830242e-06, "tIoU": 0.5277689933776856, "tRecall_50": 0.55625, "tRecall_75": 0.3625, "tRecall_90": 0.25, "token_acc": 0.9389479422185881, "epoch": 0.7437468941527249, "step": 2245}, {"loss": 0.15498046875, "grad_norm": 0.5734337568283081, "learning_rate": 1.607572365057934e-06, "tIoU": 0.5635070323944091, "tRecall_50": 0.625, "tRecall_75": 0.475, "tRecall_90": 0.2875, "token_acc": 0.9411444141689373, "epoch": 0.7454033460327978, "step": 2250}, {"loss": 0.148138427734375, "grad_norm": 0.5429418683052063, "learning_rate": 1.5879162121991614e-06, "tIoU": 0.5375117301940918, "tRecall_50": 0.575, "tRecall_75": 0.40625, "tRecall_90": 0.24375, "token_acc": 0.9419143714207799, "epoch": 0.7470597979128706, "step": 2255}, {"loss": 0.1737548828125, "grad_norm": 0.4823099374771118, "learning_rate": 1.5683582604183124e-06, "tIoU": 0.4863481044769287, "tRecall_50": 0.51875, "tRecall_75": 0.3375, "tRecall_90": 0.19375, "token_acc": 0.9344262295081968, "epoch": 0.7487162497929435, "step": 2260}, {"loss": 0.159820556640625, "grad_norm": 0.576435387134552, "learning_rate": 1.5488990726007297e-06, "tIoU": 0.5188336849212647, "tRecall_50": 0.5375, "tRecall_75": 0.33125, "tRecall_90": 0.25, "token_acc": 0.9416575790621592, "epoch": 0.7503727016730164, "step": 2265}, {"loss": 0.1625244140625, "grad_norm": 0.7238029837608337, "learning_rate": 1.5295392087892884e-06, "tIoU": 0.5586510181427002, "tRecall_50": 0.575, "tRecall_75": 0.36875, "tRecall_90": 0.25625, "token_acc": 0.9395008138903961, "epoch": 0.7520291535530893, "step": 2270}, {"loss": 0.16297607421875, "grad_norm": 0.6263623833656311, "learning_rate": 1.5102792261682813e-06, "tIoU": 0.536182975769043, "tRecall_50": 0.575, "tRecall_75": 0.375, "tRecall_90": 0.26875, "token_acc": 0.9403107113654947, "epoch": 0.7536856054331622, "step": 2275}, {"loss": 0.17288818359375, "grad_norm": 0.6171078681945801, "learning_rate": 1.4911196790473876e-06, "tIoU": 0.510650634765625, "tRecall_50": 0.525, "tRecall_75": 0.3875, "tRecall_90": 0.25625, "token_acc": 0.9331340038053819, "epoch": 0.755342057313235, "step": 2280}, {"loss": 0.15936279296875, "grad_norm": 0.5526485443115234, "learning_rate": 1.472061118845703e-06, "tIoU": 0.5104878902435303, "tRecall_50": 0.525, "tRecall_75": 0.3875, "tRecall_90": 0.23125, "token_acc": 0.9381639880141651, "epoch": 0.7569985091933079, "step": 2285}, {"loss": 0.14937744140625, "grad_norm": 0.7188597321510315, "learning_rate": 1.4531040940758922e-06, "tIoU": 0.5513885498046875, "tRecall_50": 0.63125, "tRecall_75": 0.4125, "tRecall_90": 0.2625, "token_acc": 0.9457553566585299, "epoch": 0.7586549610733808, "step": 2290}, {"loss": 0.16483154296875, "grad_norm": 0.5185531973838806, "learning_rate": 1.4342491503283862e-06, "tIoU": 0.5377781391143799, "tRecall_50": 0.5875, "tRecall_75": 0.43125, "tRecall_90": 0.275, "token_acc": 0.9410474362489718, "epoch": 0.7603114129534537, "step": 2295}, {"loss": 0.156219482421875, "grad_norm": 1.1246395111083984, "learning_rate": 1.415496830255685e-06, "tIoU": 0.5359954833984375, "tRecall_50": 0.59375, "tRecall_75": 0.3875, "tRecall_90": 0.26875, "token_acc": 0.9422185881711638, "epoch": 0.7619678648335266, "step": 2300}, {"loss": 0.16826171875, "grad_norm": 0.6118795275688171, "learning_rate": 1.3968476735567392e-06, "tIoU": 0.5206178665161133, "tRecall_50": 0.5625, "tRecall_75": 0.3875, "tRecall_90": 0.2625, "token_acc": 0.9380965366784838, "epoch": 0.7636243167135994, "step": 2305}, {"loss": 0.1671875, "grad_norm": 0.6857731342315674, "learning_rate": 1.378302216961423e-06, "tIoU": 0.5181430816650391, "tRecall_50": 0.55625, "tRecall_75": 0.35625, "tRecall_90": 0.2125, "token_acc": 0.9378055404671374, "epoch": 0.7652807685936723, "step": 2310}, {"loss": 0.15540771484375, "grad_norm": 0.5392675995826721, "learning_rate": 1.3598609942150765e-06, "tIoU": 0.547668170928955, "tRecall_50": 0.6125, "tRecall_75": 0.3875, "tRecall_90": 0.24375, "token_acc": 0.9385245901639344, "epoch": 0.7669372204737452, "step": 2315}, {"loss": 0.167828369140625, "grad_norm": 0.6362417936325073, "learning_rate": 1.3415245360631519e-06, "tIoU": 0.5249619960784913, "tRecall_50": 0.56875, "tRecall_75": 0.39375, "tRecall_90": 0.225, "token_acc": 0.9399782726778925, "epoch": 0.7685936723538181, "step": 2320}, {"loss": 0.1555908203125, "grad_norm": 0.7602875232696533, "learning_rate": 1.3232933702359357e-06, "tIoU": 0.5604390621185302, "tRecall_50": 0.6125, "tRecall_75": 0.40625, "tRecall_90": 0.2875, "token_acc": 0.9440959912735206, "epoch": 0.770250124233891, "step": 2325}, {"loss": 0.15029296875, "grad_norm": 0.6373392939567566, "learning_rate": 1.30516802143336e-06, "tIoU": 0.5182899951934814, "tRecall_50": 0.55, "tRecall_75": 0.38125, "tRecall_90": 0.2375, "token_acc": 0.9408235614944096, "epoch": 0.7719065761139638, "step": 2330}, {"loss": 0.1555419921875, "grad_norm": 0.7477315068244934, "learning_rate": 1.2871490113099066e-06, "tIoU": 0.5171179294586181, "tRecall_50": 0.53125, "tRecall_75": 0.35, "tRecall_90": 0.20625, "token_acc": 0.9402660874287265, "epoch": 0.7735630279940368, "step": 2335}, {"loss": 0.15079345703125, "grad_norm": 0.789494514465332, "learning_rate": 1.269236858459585e-06, "tIoU": 0.5262135982513427, "tRecall_50": 0.56875, "tRecall_75": 0.36875, "tRecall_90": 0.2375, "token_acc": 0.9411603721948549, "epoch": 0.7752194798741097, "step": 2340}, {"loss": 0.15489501953125, "grad_norm": 0.5260175466537476, "learning_rate": 1.251432078401011e-06, "tIoU": 0.5641520500183106, "tRecall_50": 0.61875, "tRecall_75": 0.3875, "tRecall_90": 0.25625, "token_acc": 0.9441431670281996, "epoch": 0.7768759317541826, "step": 2345}, {"loss": 0.16513671875, "grad_norm": 0.6776566505432129, "learning_rate": 1.2337351835625772e-06, "tIoU": 0.5146319389343261, "tRecall_50": 0.575, "tRecall_75": 0.4, "tRecall_90": 0.2375, "token_acc": 0.9373461012311901, "epoch": 0.7785323836342555, "step": 2350}, {"loss": 0.16092529296875, "grad_norm": 0.7546380162239075, "learning_rate": 1.2161466832676888e-06, "tIoU": 0.5249964714050293, "tRecall_50": 0.5625, "tRecall_75": 0.40625, "tRecall_90": 0.23125, "token_acc": 0.941304347826087, "epoch": 0.7801888355143283, "step": 2355}, {"loss": 0.14600830078125, "grad_norm": 0.5215368866920471, "learning_rate": 1.1986670837201186e-06, "tIoU": 0.5563305854797364, "tRecall_50": 0.6, "tRecall_75": 0.40625, "tRecall_90": 0.3, "token_acc": 0.9457617879531207, "epoch": 0.7818452873944012, "step": 2360}, {"loss": 0.172119140625, "grad_norm": 0.7974917888641357, "learning_rate": 1.1812968879894387e-06, "tIoU": 0.5147394657135009, "tRecall_50": 0.575, "tRecall_75": 0.34375, "tRecall_90": 0.1625, "token_acc": 0.9330976339407125, "epoch": 0.7835017392744741, "step": 2365}, {"loss": 0.16275634765625, "grad_norm": 0.39690589904785156, "learning_rate": 1.1640365959965305e-06, "tIoU": 0.5097333431243897, "tRecall_50": 0.58125, "tRecall_75": 0.35, "tRecall_90": 0.2125, "token_acc": 0.9370420624151967, "epoch": 0.785158191154547, "step": 2370}, {"loss": 0.15360107421875, "grad_norm": 0.48728054761886597, "learning_rate": 1.1468867044992065e-06, "tIoU": 0.548112154006958, "tRecall_50": 0.575, "tRecall_75": 0.4, "tRecall_90": 0.28125, "token_acc": 0.9423024336888159, "epoch": 0.7868146430346199, "step": 2375}, {"loss": 0.153314208984375, "grad_norm": 0.642907977104187, "learning_rate": 1.1298477070779123e-06, "tIoU": 0.5524973392486572, "tRecall_50": 0.59375, "tRecall_75": 0.4125, "tRecall_90": 0.26875, "token_acc": 0.9427408412483039, "epoch": 0.7884710949146927, "step": 2380}, {"loss": 0.1518798828125, "grad_norm": 0.4547934830188751, "learning_rate": 1.1129200941215185e-06, "tIoU": 0.5372862339019775, "tRecall_50": 0.5875, "tRecall_75": 0.39375, "tRecall_90": 0.2375, "token_acc": 0.9394188596491229, "epoch": 0.7901275467947656, "step": 2385}, {"loss": 0.16217041015625, "grad_norm": 0.45014676451683044, "learning_rate": 1.096104352813206e-06, "tIoU": 0.5642598152160645, "tRecall_50": 0.63125, "tRecall_75": 0.41875, "tRecall_90": 0.3125, "token_acc": 0.9434117003827228, "epoch": 0.7917839986748385, "step": 2390}, {"loss": 0.1718505859375, "grad_norm": 0.584336519241333, "learning_rate": 1.0794009671164484e-06, "tIoU": 0.503229808807373, "tRecall_50": 0.56875, "tRecall_75": 0.3125, "tRecall_90": 0.2, "token_acc": 0.9349371241115364, "epoch": 0.7934404505549114, "step": 2395}, {"loss": 0.15487060546875, "grad_norm": 0.7591623663902283, "learning_rate": 1.062810417761082e-06, "tIoU": 0.5614675521850586, "tRecall_50": 0.6, "tRecall_75": 0.41875, "tRecall_90": 0.3, "token_acc": 0.9457322061630761, "epoch": 0.7950969024349842, "step": 2400}, {"loss": 0.16737060546875, "grad_norm": 0.6303367018699646, "learning_rate": 1.046333182229472e-06, "tIoU": 0.4994485855102539, "tRecall_50": 0.55625, "tRecall_75": 0.3375, "tRecall_90": 0.21875, "token_acc": 0.9360892031547458, "epoch": 0.7967533543150571, "step": 2405}, {"loss": 0.1697509765625, "grad_norm": 0.8385005593299866, "learning_rate": 1.0299697347427645e-06, "tIoU": 0.554076862335205, "tRecall_50": 0.625, "tRecall_75": 0.38125, "tRecall_90": 0.25, "token_acc": 0.9413689664575947, "epoch": 0.79840980619513, "step": 2410}, {"loss": 0.14515380859375, "grad_norm": 0.6249813437461853, "learning_rate": 1.0137205462472427e-06, "tIoU": 0.5932907581329345, "tRecall_50": 0.66875, "tRecall_75": 0.4625, "tRecall_90": 0.28125, "token_acc": 0.9468638729115311, "epoch": 0.8000662580752029, "step": 2415}, {"loss": 0.16220703125, "grad_norm": 2.346714973449707, "learning_rate": 9.975860844007774e-07, "tIoU": 0.47458863258361816, "tRecall_50": 0.51875, "tRecall_75": 0.31875, "tRecall_90": 0.20625, "token_acc": 0.9378068739770867, "epoch": 0.8017227099552758, "step": 2420}, {"loss": 0.1552490234375, "grad_norm": 0.6921612620353699, "learning_rate": 9.815668135593548e-07, "tIoU": 0.5383291244506836, "tRecall_50": 0.54375, "tRecall_75": 0.4, "tRecall_90": 0.3, "token_acc": 0.9442622950819672, "epoch": 0.8033791618353486, "step": 2425}, {"loss": 0.15458984375, "grad_norm": 0.7046741843223572, "learning_rate": 9.656631947637235e-07, "tIoU": 0.5690701007843018, "tRecall_50": 0.60625, "tRecall_75": 0.45625, "tRecall_90": 0.25, "token_acc": 0.9448669201520913, "epoch": 0.8050356137154215, "step": 2430}, {"loss": 0.15469970703125, "grad_norm": 0.5167131423950195, "learning_rate": 9.498756857261243e-07, "tIoU": 0.5229488849639893, "tRecall_50": 0.5625, "tRecall_75": 0.4, "tRecall_90": 0.26875, "token_acc": 0.9429273620972146, "epoch": 0.8066920655954944, "step": 2435}, {"loss": 0.167047119140625, "grad_norm": 0.472789466381073, "learning_rate": 9.342047408171117e-07, "tIoU": 0.5211413860321045, "tRecall_50": 0.56875, "tRecall_75": 0.4, "tRecall_90": 0.25, "token_acc": 0.9389146441232615, "epoch": 0.8083485174755674, "step": 2440}, {"loss": 0.15396728515625, "grad_norm": 0.4715271294116974, "learning_rate": 9.186508110524795e-07, "tIoU": 0.5372179031372071, "tRecall_50": 0.56875, "tRecall_75": 0.425, "tRecall_90": 0.26875, "token_acc": 0.942888522789676, "epoch": 0.8100049693556403, "step": 2445}, {"loss": 0.155706787109375, "grad_norm": 0.5346218347549438, "learning_rate": 9.032143440802837e-07, "tIoU": 0.5261980533599854, "tRecall_50": 0.5625, "tRecall_75": 0.40625, "tRecall_90": 0.28125, "token_acc": 0.941464742717125, "epoch": 0.811661421235713, "step": 2450}, {"loss": 0.15142822265625, "grad_norm": 0.5118564367294312, "learning_rate": 8.878957841679542e-07, "tIoU": 0.5573920249938965, "tRecall_50": 0.625, "tRecall_75": 0.4125, "tRecall_90": 0.2625, "token_acc": 0.9446103719793646, "epoch": 0.813317873115786, "step": 2455}, {"loss": 0.15771484375, "grad_norm": 0.5679037570953369, "learning_rate": 8.726955721895103e-07, "tIoU": 0.5401461601257325, "tRecall_50": 0.5875, "tRecall_75": 0.425, "tRecall_90": 0.2875, "token_acc": 0.9441707717569786, "epoch": 0.8149743249958589, "step": 2460}, {"loss": 0.16822509765625, "grad_norm": 0.5475548505783081, "learning_rate": 8.576141456128734e-07, "tIoU": 0.4817670345306396, "tRecall_50": 0.5, "tRecall_75": 0.35, "tRecall_90": 0.225, "token_acc": 0.9360830374214696, "epoch": 0.8166307768759318, "step": 2465}, {"loss": 0.1654541015625, "grad_norm": 0.49165377020835876, "learning_rate": 8.426519384872733e-07, "tIoU": 0.5372585773468017, "tRecall_50": 0.59375, "tRecall_75": 0.41875, "tRecall_90": 0.2625, "token_acc": 0.9378762999452654, "epoch": 0.8182872287560047, "step": 2470}, {"loss": 0.1543701171875, "grad_norm": 0.6567497253417969, "learning_rate": 8.278093814307637e-07, "tIoU": 0.5190737247467041, "tRecall_50": 0.56875, "tRecall_75": 0.4125, "tRecall_90": 0.225, "token_acc": 0.9400656814449918, "epoch": 0.8199436806360775, "step": 2475}, {"loss": 0.15341796875, "grad_norm": 0.6542906761169434, "learning_rate": 8.130869016178178e-07, "tIoU": 0.5407592296600342, "tRecall_50": 0.60625, "tRecall_75": 0.38125, "tRecall_90": 0.2625, "token_acc": 0.9434889434889435, "epoch": 0.8216001325161504, "step": 2480}, {"loss": 0.162042236328125, "grad_norm": 0.6917836666107178, "learning_rate": 7.984849227670421e-07, "tIoU": 0.5039559364318847, "tRecall_50": 0.53125, "tRecall_75": 0.3375, "tRecall_90": 0.2375, "token_acc": 0.9401455918037207, "epoch": 0.8232565843962233, "step": 2485}, {"loss": 0.17220458984375, "grad_norm": 0.6990655660629272, "learning_rate": 7.840038651289794e-07, "tIoU": 0.5036819458007813, "tRecall_50": 0.5375, "tRecall_75": 0.35, "tRecall_90": 0.25625, "token_acc": 0.9378898833740168, "epoch": 0.8249130362762962, "step": 2490}, {"loss": 0.161029052734375, "grad_norm": 0.6285939812660217, "learning_rate": 7.696441454740134e-07, "tIoU": 0.5125839710235596, "tRecall_50": 0.55, "tRecall_75": 0.33125, "tRecall_90": 0.23125, "token_acc": 0.9377559377559378, "epoch": 0.8265694881563691, "step": 2495}, {"loss": 0.16112060546875, "grad_norm": 0.6815149188041687, "learning_rate": 7.554061770803733e-07, "tIoU": 0.5246678352355957, "tRecall_50": 0.58125, "tRecall_75": 0.375, "tRecall_90": 0.21875, "token_acc": 0.9360260233125508, "epoch": 0.8282259400364419, "step": 2500}, {"loss": 0.1623291015625, "grad_norm": 0.7072858810424805, "learning_rate": 7.412903697222429e-07, "tIoU": 0.5076375961303711, "tRecall_50": 0.5125, "tRecall_75": 0.38125, "tRecall_90": 0.225, "token_acc": 0.9397163120567376, "epoch": 0.8298823919165148, "step": 2505}, {"loss": 0.15618896484375, "grad_norm": 0.5564939379692078, "learning_rate": 7.272971296579629e-07, "tIoU": 0.5498382568359375, "tRecall_50": 0.6, "tRecall_75": 0.4125, "tRecall_90": 0.21875, "token_acc": 0.9418732782369146, "epoch": 0.8315388437965877, "step": 2510}, {"loss": 0.15855712890625, "grad_norm": 0.5424681901931763, "learning_rate": 7.13426859618338e-07, "tIoU": 0.5336521625518799, "tRecall_50": 0.58125, "tRecall_75": 0.3875, "tRecall_90": 0.225, "token_acc": 0.9439277899343544, "epoch": 0.8331952956766606, "step": 2515}, {"loss": 0.165234375, "grad_norm": 0.48574844002723694, "learning_rate": 6.996799587950542e-07, "tIoU": 0.5291247367858887, "tRecall_50": 0.55, "tRecall_75": 0.40625, "tRecall_90": 0.2875, "token_acc": 0.9401639344262295, "epoch": 0.8348517475567335, "step": 2520}, {"loss": 0.153076171875, "grad_norm": 0.656525194644928, "learning_rate": 6.860568228291808e-07, "tIoU": 0.5800621509552002, "tRecall_50": 0.6375, "tRecall_75": 0.45, "tRecall_90": 0.26875, "token_acc": 0.9424440807419531, "epoch": 0.8365081994368063, "step": 2525}, {"loss": 0.15035400390625, "grad_norm": 0.7389978766441345, "learning_rate": 6.725578437997837e-07, "tIoU": 0.5640317916870117, "tRecall_50": 0.6125, "tRecall_75": 0.425, "tRecall_90": 0.2625, "token_acc": 0.9400328587075575, "epoch": 0.8381646513168792, "step": 2530}, {"loss": 0.1676513671875, "grad_norm": 0.7263797521591187, "learning_rate": 6.591834102126516e-07, "tIoU": 0.49884805679321287, "tRecall_50": 0.53125, "tRecall_75": 0.34375, "tRecall_90": 0.2375, "token_acc": 0.9385869565217392, "epoch": 0.8398211031969521, "step": 2535}, {"loss": 0.1603759765625, "grad_norm": 0.6837959885597229, "learning_rate": 6.459339069891036e-07, "tIoU": 0.5144101619720459, "tRecall_50": 0.50625, "tRecall_75": 0.3875, "tRecall_90": 0.2875, "token_acc": 0.9393442622950819, "epoch": 0.841477555077025, "step": 2540}, {"loss": 0.1583740234375, "grad_norm": 0.5565042495727539, "learning_rate": 6.328097154549146e-07, "tIoU": 0.5358321666717529, "tRecall_50": 0.56875, "tRecall_75": 0.375, "tRecall_90": 0.23125, "token_acc": 0.9405860010851872, "epoch": 0.843134006957098, "step": 2545}, {"loss": 0.17939453125, "grad_norm": 0.4988414943218231, "learning_rate": 6.198112133293438e-07, "tIoU": 0.5300264835357666, "tRecall_50": 0.5375, "tRecall_75": 0.40625, "tRecall_90": 0.28125, "token_acc": 0.9378746594005449, "epoch": 0.8447904588371707, "step": 2550}, {"loss": 0.16768798828125, "grad_norm": 0.6543470621109009, "learning_rate": 6.069387747142591e-07, "tIoU": 0.5053749084472656, "tRecall_50": 0.51875, "tRecall_75": 0.3625, "tRecall_90": 0.28125, "token_acc": 0.93435031326614, "epoch": 0.8464469107172437, "step": 2555}, {"loss": 0.16727294921875, "grad_norm": 0.9791786074638367, "learning_rate": 5.941927700833727e-07, "tIoU": 0.5062736988067627, "tRecall_50": 0.55625, "tRecall_75": 0.29375, "tRecall_90": 0.2125, "token_acc": 0.937363238512035, "epoch": 0.8481033625973166, "step": 2560}, {"loss": 0.164141845703125, "grad_norm": 0.6862460374832153, "learning_rate": 5.815735662715782e-07, "tIoU": 0.5086766719818115, "tRecall_50": 0.54375, "tRecall_75": 0.375, "tRecall_90": 0.25625, "token_acc": 0.9397491821155943, "epoch": 0.8497598144773895, "step": 2565}, {"loss": 0.14600830078125, "grad_norm": 0.5261525511741638, "learning_rate": 5.690815264643929e-07, "tIoU": 0.5660731315612793, "tRecall_50": 0.6, "tRecall_75": 0.41875, "tRecall_90": 0.2625, "token_acc": 0.9455189321710705, "epoch": 0.8514162663574624, "step": 2570}, {"loss": 0.16983642578125, "grad_norm": 0.6874206066131592, "learning_rate": 5.567170101875074e-07, "tIoU": 0.5341813087463378, "tRecall_50": 0.5625, "tRecall_75": 0.40625, "tRecall_90": 0.2875, "token_acc": 0.9367396593673966, "epoch": 0.8530727182375352, "step": 2575}, {"loss": 0.16873779296875, "grad_norm": 0.5525768399238586, "learning_rate": 5.444803732964333e-07, "tIoU": 0.4944866180419922, "tRecall_50": 0.5625, "tRecall_75": 0.33125, "tRecall_90": 0.16875, "token_acc": 0.937244201909959, "epoch": 0.8547291701176081, "step": 2580}, {"loss": 0.172900390625, "grad_norm": 0.5512920618057251, "learning_rate": 5.323719679662647e-07, "tIoU": 0.5066828250885009, "tRecall_50": 0.55625, "tRecall_75": 0.35625, "tRecall_90": 0.225, "token_acc": 0.9372784292337061, "epoch": 0.856385621997681, "step": 2585}, {"loss": 0.174609375, "grad_norm": 1.3920177221298218, "learning_rate": 5.203921426815467e-07, "tIoU": 0.528574800491333, "tRecall_50": 0.56875, "tRecall_75": 0.3875, "tRecall_90": 0.26875, "token_acc": 0.9373803664205633, "epoch": 0.8580420738777539, "step": 2590}, {"loss": 0.17049560546875, "grad_norm": 0.6287409663200378, "learning_rate": 5.085412422262364e-07, "tIoU": 0.4851510047912598, "tRecall_50": 0.51875, "tRecall_75": 0.33125, "tRecall_90": 0.1625, "token_acc": 0.931588988825293, "epoch": 0.8596985257578267, "step": 2595}, {"loss": 0.152935791015625, "grad_norm": 0.5815898776054382, "learning_rate": 4.968196076737847e-07, "tIoU": 0.5580605030059814, "tRecall_50": 0.625, "tRecall_75": 0.39375, "tRecall_90": 0.24375, "token_acc": 0.9416278432447246, "epoch": 0.8613549776378996, "step": 2600}, {"loss": 0.15257568359375, "grad_norm": 0.667481541633606, "learning_rate": 4.852275763773251e-07, "tIoU": 0.5585962295532226, "tRecall_50": 0.61875, "tRecall_75": 0.375, "tRecall_90": 0.2375, "token_acc": 0.9419302071973827, "epoch": 0.8630114295179725, "step": 2605}, {"loss": 0.1552734375, "grad_norm": 0.5576724410057068, "learning_rate": 4.737654819599563e-07, "tIoU": 0.5506963729858398, "tRecall_50": 0.60625, "tRecall_75": 0.4125, "tRecall_90": 0.28125, "token_acc": 0.9373798407031035, "epoch": 0.8646678813980454, "step": 2610}, {"loss": 0.1733154296875, "grad_norm": 0.45226308703422546, "learning_rate": 4.624336543051433e-07, "tIoU": 0.4877234935760498, "tRecall_50": 0.53125, "tRecall_75": 0.35, "tRecall_90": 0.175, "token_acc": 0.9375338020551649, "epoch": 0.8663243332781183, "step": 2615}, {"loss": 0.154150390625, "grad_norm": 0.757668137550354, "learning_rate": 4.5123241954722673e-07, "tIoU": 0.5472731113433837, "tRecall_50": 0.575, "tRecall_75": 0.41875, "tRecall_90": 0.28125, "token_acc": 0.9441128594682583, "epoch": 0.8679807851581911, "step": 2620}, {"loss": 0.1505615234375, "grad_norm": 0.6104984879493713, "learning_rate": 4.4016210006203096e-07, "tIoU": 0.5675601005554199, "tRecall_50": 0.56875, "tRecall_75": 0.45625, "tRecall_90": 0.30625, "token_acc": 0.9440959912735206, "epoch": 0.869637237038264, "step": 2625}, {"loss": 0.15528564453125, "grad_norm": 0.6254884004592896, "learning_rate": 4.292230144575904e-07, "tIoU": 0.5098196029663086, "tRecall_50": 0.5375, "tRecall_75": 0.40625, "tRecall_90": 0.225, "token_acc": 0.9385034013605442, "epoch": 0.8712936889183369, "step": 2630}, {"loss": 0.164764404296875, "grad_norm": 0.8128376603126526, "learning_rate": 4.184154775649768e-07, "tIoU": 0.5138954162597656, "tRecall_50": 0.5625, "tRecall_75": 0.4, "tRecall_90": 0.20625, "token_acc": 0.9365986394557824, "epoch": 0.8729501407984098, "step": 2635}, {"loss": 0.148931884765625, "grad_norm": 0.5039224028587341, "learning_rate": 4.0773980042923986e-07, "tIoU": 0.5136477947235107, "tRecall_50": 0.53125, "tRecall_75": 0.3625, "tRecall_90": 0.25, "token_acc": 0.9414951245937161, "epoch": 0.8746065926784827, "step": 2640}, {"loss": 0.17333984375, "grad_norm": 0.4894895553588867, "learning_rate": 3.9719629030045616e-07, "tIoU": 0.469511604309082, "tRecall_50": 0.4625, "tRecall_75": 0.3125, "tRecall_90": 0.18125, "token_acc": 0.9365468409586056, "epoch": 0.8762630445585555, "step": 2645}, {"loss": 0.16063232421875, "grad_norm": 0.4093848168849945, "learning_rate": 3.867852506248843e-07, "tIoU": 0.49820685386657715, "tRecall_50": 0.5375, "tRecall_75": 0.35625, "tRecall_90": 0.24375, "token_acc": 0.9370552818828681, "epoch": 0.8779194964386284, "step": 2650}, {"loss": 0.166009521484375, "grad_norm": 0.5204543471336365, "learning_rate": 3.765069810362326e-07, "tIoU": 0.5327613353729248, "tRecall_50": 0.5625, "tRecall_75": 0.3875, "tRecall_90": 0.24375, "token_acc": 0.9399890889252591, "epoch": 0.8795759483187013, "step": 2655}, {"loss": 0.15164794921875, "grad_norm": 0.6944509148597717, "learning_rate": 3.6636177734703405e-07, "tIoU": 0.5492005825042725, "tRecall_50": 0.60625, "tRecall_75": 0.40625, "tRecall_90": 0.25, "token_acc": 0.9418003807451727, "epoch": 0.8812324001987742, "step": 2660}, {"loss": 0.145611572265625, "grad_norm": 0.7271901965141296, "learning_rate": 3.563499315401386e-07, "tIoU": 0.5534201622009277, "tRecall_50": 0.6125, "tRecall_75": 0.425, "tRecall_90": 0.30625, "token_acc": 0.9439608269858542, "epoch": 0.8828888520788472, "step": 2665}, {"loss": 0.1537109375, "grad_norm": 0.6131239533424377, "learning_rate": 3.4647173176029835e-07, "tIoU": 0.5357179164886474, "tRecall_50": 0.575, "tRecall_75": 0.4125, "tRecall_90": 0.26875, "token_acc": 0.9426162632580908, "epoch": 0.88454530395892, "step": 2670}, {"loss": 0.15107421875, "grad_norm": 0.49681997299194336, "learning_rate": 3.3672746230588803e-07, "tIoU": 0.576555585861206, "tRecall_50": 0.63125, "tRecall_75": 0.43125, "tRecall_90": 0.25625, "token_acc": 0.9429192715411797, "epoch": 0.8862017558389929, "step": 2675}, {"loss": 0.158447265625, "grad_norm": 0.5076614618301392, "learning_rate": 3.271174036207131e-07, "tIoU": 0.5341426849365234, "tRecall_50": 0.55, "tRecall_75": 0.375, "tRecall_90": 0.25625, "token_acc": 0.9407125373946151, "epoch": 0.8878582077190658, "step": 2680}, {"loss": 0.1483642578125, "grad_norm": 0.6126746535301208, "learning_rate": 3.176418322859415e-07, "tIoU": 0.5753330230712891, "tRecall_50": 0.65, "tRecall_75": 0.45625, "tRecall_90": 0.3, "token_acc": 0.941240776168352, "epoch": 0.8895146595991387, "step": 2685}, {"loss": 0.177294921875, "grad_norm": 0.7555765509605408, "learning_rate": 3.0830102101214563e-07, "tIoU": 0.46550650596618653, "tRecall_50": 0.48125, "tRecall_75": 0.34375, "tRecall_90": 0.2125, "token_acc": 0.9303780255643187, "epoch": 0.8911711114792116, "step": 2690}, {"loss": 0.143023681640625, "grad_norm": 0.7692764401435852, "learning_rate": 2.990952386314505e-07, "tIoU": 0.5630221366882324, "tRecall_50": 0.625, "tRecall_75": 0.425, "tRecall_90": 0.3, "token_acc": 0.944505194095134, "epoch": 0.8928275633592844, "step": 2695}, {"loss": 0.146484375, "grad_norm": 0.5720428228378296, "learning_rate": 2.900247500897974e-07, "tIoU": 0.5490079879760742, "tRecall_50": 0.6, "tRecall_75": 0.4125, "tRecall_90": 0.275, "token_acc": 0.9426386233269598, "epoch": 0.8944840152393573, "step": 2700}, {"loss": 0.152166748046875, "grad_norm": 0.8671318292617798, "learning_rate": 2.8108981643932e-07, "tIoU": 0.571419620513916, "tRecall_50": 0.63125, "tRecall_75": 0.41875, "tRecall_90": 0.23125, "token_acc": 0.9429041998353006, "epoch": 0.8961404671194302, "step": 2705}, {"loss": 0.147705078125, "grad_norm": 0.6312313079833984, "learning_rate": 2.7229069483083015e-07, "tIoU": 0.569254732131958, "tRecall_50": 0.61875, "tRecall_75": 0.43125, "tRecall_90": 0.26875, "token_acc": 0.9452316076294278, "epoch": 0.8977969189995031, "step": 2710}, {"loss": 0.15697021484375, "grad_norm": 0.9674606919288635, "learning_rate": 2.636276385064157e-07, "tIoU": 0.5186336040496826, "tRecall_50": 0.575, "tRecall_75": 0.375, "tRecall_90": 0.23125, "token_acc": 0.9401360544217687, "epoch": 0.899453370879576, "step": 2715}, {"loss": 0.15101318359375, "grad_norm": 0.5326899290084839, "learning_rate": 2.5510089679215646e-07, "tIoU": 0.560830020904541, "tRecall_50": 0.5875, "tRecall_75": 0.45625, "tRecall_90": 0.3, "token_acc": 0.9469074986316366, "epoch": 0.9011098227596488, "step": 2720}, {"loss": 0.16943359375, "grad_norm": 0.510621964931488, "learning_rate": 2.4671071509094367e-07, "tIoU": 0.489136266708374, "tRecall_50": 0.53125, "tRecall_75": 0.36875, "tRecall_90": 0.19375, "token_acc": 0.9332066250339397, "epoch": 0.9027662746397217, "step": 2725}, {"loss": 0.169390869140625, "grad_norm": 0.6251363158226013, "learning_rate": 2.3845733487541824e-07, "tIoU": 0.4824803829193115, "tRecall_50": 0.53125, "tRecall_75": 0.31875, "tRecall_90": 0.19375, "token_acc": 0.9344083082809511, "epoch": 0.9044227265197946, "step": 2730}, {"loss": 0.1482421875, "grad_norm": 0.6128382086753845, "learning_rate": 2.3034099368102525e-07, "tIoU": 0.5446168899536132, "tRecall_50": 0.575, "tRecall_75": 0.41875, "tRecall_90": 0.2875, "token_acc": 0.9440825190010858, "epoch": 0.9060791783998675, "step": 2735}, {"loss": 0.159869384765625, "grad_norm": 1.0730916261672974, "learning_rate": 2.2236192509916933e-07, "tIoU": 0.4948151111602783, "tRecall_50": 0.5375, "tRecall_75": 0.3375, "tRecall_90": 0.25, "token_acc": 0.9408893709327549, "epoch": 0.9077356302799404, "step": 2740}, {"loss": 0.15233154296875, "grad_norm": 0.5609004497528076, "learning_rate": 2.1452035877050137e-07, "tIoU": 0.5627793312072754, "tRecall_50": 0.60625, "tRecall_75": 0.425, "tRecall_90": 0.325, "token_acc": 0.9446413962367057, "epoch": 0.9093920821600132, "step": 2745}, {"loss": 0.153057861328125, "grad_norm": 0.6590545773506165, "learning_rate": 2.068165203783029e-07, "tIoU": 0.5494290351867676, "tRecall_50": 0.63125, "tRecall_75": 0.35625, "tRecall_90": 0.275, "token_acc": 0.9417052574230454, "epoch": 0.9110485340400861, "step": 2750}, {"loss": 0.1706298828125, "grad_norm": 0.4576837122440338, "learning_rate": 1.992506316419912e-07, "tIoU": 0.5233198642730713, "tRecall_50": 0.55, "tRecall_75": 0.39375, "tRecall_90": 0.23125, "token_acc": 0.934640522875817, "epoch": 0.912704985920159, "step": 2755}, {"loss": 0.1559814453125, "grad_norm": 0.6967729330062866, "learning_rate": 1.9182291031074275e-07, "tIoU": 0.5519247055053711, "tRecall_50": 0.6, "tRecall_75": 0.4125, "tRecall_90": 0.24375, "token_acc": 0.9383783783783783, "epoch": 0.9143614378002319, "step": 2760}, {"loss": 0.17286376953125, "grad_norm": 0.4412071406841278, "learning_rate": 1.8453357015722097e-07, "tIoU": 0.47763400077819823, "tRecall_50": 0.5125, "tRecall_75": 0.34375, "tRecall_90": 0.225, "token_acc": 0.9332971800433839, "epoch": 0.9160178896803048, "step": 2765}, {"loss": 0.15433349609375, "grad_norm": 0.4622722864151001, "learning_rate": 1.7738282097142622e-07, "tIoU": 0.5353363513946533, "tRecall_50": 0.575, "tRecall_75": 0.4, "tRecall_90": 0.24375, "token_acc": 0.9383319750067917, "epoch": 0.9176743415603776, "step": 2770}, {"loss": 0.1578369140625, "grad_norm": 0.7105121612548828, "learning_rate": 1.7037086855465902e-07, "tIoU": 0.5579864025115967, "tRecall_50": 0.625, "tRecall_75": 0.425, "tRecall_90": 0.275, "token_acc": 0.9409513395297977, "epoch": 0.9193307934404505, "step": 2775}, {"loss": 0.18677978515625, "grad_norm": 0.516262948513031, "learning_rate": 1.63497914713594e-07, "tIoU": 0.4627009391784668, "tRecall_50": 0.48125, "tRecall_75": 0.35625, "tRecall_90": 0.2375, "token_acc": 0.93, "epoch": 0.9209872453205235, "step": 2780}, {"loss": 0.15511474609375, "grad_norm": 0.5634688138961792, "learning_rate": 1.5676415725447424e-07, "tIoU": 0.5456976890563965, "tRecall_50": 0.6, "tRecall_75": 0.44375, "tRecall_90": 0.2625, "token_acc": 0.9393118396098619, "epoch": 0.9226436972005964, "step": 2785}, {"loss": 0.160699462890625, "grad_norm": 0.648868978023529, "learning_rate": 1.5016978997741903e-07, "tIoU": 0.5549224853515625, "tRecall_50": 0.6, "tRecall_75": 0.41875, "tRecall_90": 0.2625, "token_acc": 0.9421239421239421, "epoch": 0.9243001490806692, "step": 2790}, {"loss": 0.1560546875, "grad_norm": 0.5581037998199463, "learning_rate": 1.4371500267084337e-07, "tIoU": 0.5302978038787842, "tRecall_50": 0.5875, "tRecall_75": 0.3875, "tRecall_90": 0.25, "token_acc": 0.9379781420765028, "epoch": 0.9259566009607421, "step": 2795}, {"loss": 0.16649169921875, "grad_norm": 1.205271601676941, "learning_rate": 1.3739998110599618e-07, "tIoU": 0.5526116847991943, "tRecall_50": 0.60625, "tRecall_75": 0.4, "tRecall_90": 0.24375, "token_acc": 0.9417186229330442, "epoch": 0.927613052840815, "step": 2800}, {"loss": 0.14595947265625, "grad_norm": 0.5804457068443298, "learning_rate": 1.3122490703161793e-07, "tIoU": 0.5765332221984864, "tRecall_50": 0.61875, "tRecall_75": 0.4375, "tRecall_90": 0.325, "token_acc": 0.9430982847808331, "epoch": 0.9292695047208879, "step": 2805}, {"loss": 0.16298828125, "grad_norm": 0.5786797404289246, "learning_rate": 1.2518995816870327e-07, "tIoU": 0.5211046218872071, "tRecall_50": 0.5875, "tRecall_75": 0.35, "tRecall_90": 0.23125, "token_acc": 0.9343995662781241, "epoch": 0.9309259566009608, "step": 2810}, {"loss": 0.151806640625, "grad_norm": 0.731719970703125, "learning_rate": 1.192953082053927e-07, "tIoU": 0.5116126060485839, "tRecall_50": 0.54375, "tRecall_75": 0.38125, "tRecall_90": 0.28125, "token_acc": 0.9389208436044919, "epoch": 0.9325824084810336, "step": 2815}, {"loss": 0.1593505859375, "grad_norm": 0.5219873785972595, "learning_rate": 1.1354112679196894e-07, "tIoU": 0.5594928741455079, "tRecall_50": 0.5875, "tRecall_75": 0.41875, "tRecall_90": 0.3125, "token_acc": 0.9441280173582859, "epoch": 0.9342388603611065, "step": 2820}, {"loss": 0.161572265625, "grad_norm": 0.6567935347557068, "learning_rate": 1.0792757953597522e-07, "tIoU": 0.5246160507202149, "tRecall_50": 0.6, "tRecall_75": 0.3875, "tRecall_90": 0.225, "token_acc": 0.9388706140350878, "epoch": 0.9358953122411794, "step": 2825}, {"loss": 0.15240478515625, "grad_norm": 0.6812028288841248, "learning_rate": 1.0245482799745243e-07, "tIoU": 0.5628941535949707, "tRecall_50": 0.58125, "tRecall_75": 0.425, "tRecall_90": 0.29375, "token_acc": 0.9423969423969424, "epoch": 0.9375517641212523, "step": 2830}, {"loss": 0.16417236328125, "grad_norm": 0.49755436182022095, "learning_rate": 9.712302968428567e-08, "tIoU": 0.5263060092926025, "tRecall_50": 0.58125, "tRecall_75": 0.375, "tRecall_90": 0.23125, "token_acc": 0.9415300546448088, "epoch": 0.9392082160013252, "step": 2835}, {"loss": 0.156787109375, "grad_norm": 0.6163895130157471, "learning_rate": 9.193233804767054e-08, "tIoU": 0.5490593433380127, "tRecall_50": 0.5875, "tRecall_75": 0.4125, "tRecall_90": 0.2875, "token_acc": 0.9403757146746529, "epoch": 0.940864667881398, "step": 2840}, {"loss": 0.1619384765625, "grad_norm": 0.5530081987380981, "learning_rate": 8.688290247770071e-08, "tIoU": 0.5022723197937011, "tRecall_50": 0.5375, "tRecall_75": 0.36875, "tRecall_90": 0.18125, "token_acc": 0.9364303178484108, "epoch": 0.9425211197614709, "step": 2845}, {"loss": 0.1586669921875, "grad_norm": 0.672563374042511, "learning_rate": 8.197486829906564e-08, "tIoU": 0.5501746654510498, "tRecall_50": 0.61875, "tRecall_75": 0.41875, "tRecall_90": 0.18125, "token_acc": 0.9395589436427988, "epoch": 0.9441775716415438, "step": 2850}, {"loss": 0.14420166015625, "grad_norm": 0.45273101329803467, "learning_rate": 7.720837676686743e-08, "tIoU": 0.535654354095459, "tRecall_50": 0.54375, "tRecall_75": 0.44375, "tRecall_90": 0.275, "token_acc": 0.9438415626695605, "epoch": 0.9458340235216167, "step": 2855}, {"loss": 0.1594482421875, "grad_norm": 0.5784236192703247, "learning_rate": 7.258356506255781e-08, "tIoU": 0.5448553085327148, "tRecall_50": 0.58125, "tRecall_75": 0.4625, "tRecall_90": 0.3, "token_acc": 0.9401639344262295, "epoch": 0.9474904754016896, "step": 2860}, {"loss": 0.15972900390625, "grad_norm": 0.6065331697463989, "learning_rate": 6.810056628998918e-08, "tIoU": 0.49558448791503906, "tRecall_50": 0.5, "tRecall_75": 0.36875, "tRecall_90": 0.21875, "token_acc": 0.9342069342069342, "epoch": 0.9491469272817624, "step": 2865}, {"loss": 0.15238037109375, "grad_norm": 0.547734797000885, "learning_rate": 6.375950947158205e-08, "tIoU": 0.5578720092773437, "tRecall_50": 0.59375, "tRecall_75": 0.425, "tRecall_90": 0.25625, "token_acc": 0.9423182066703116, "epoch": 0.9508033791618353, "step": 2870}, {"loss": 0.15859375, "grad_norm": 0.5336986780166626, "learning_rate": 5.9560519544614725e-08, "tIoU": 0.5313952445983887, "tRecall_50": 0.575, "tRecall_75": 0.38125, "tRecall_90": 0.275, "token_acc": 0.9424283765347885, "epoch": 0.9524598310419082, "step": 2875}, {"loss": 0.148968505859375, "grad_norm": 0.5611312985420227, "learning_rate": 5.5503717357623945e-08, "tIoU": 0.5653798103332519, "tRecall_50": 0.61875, "tRecall_75": 0.46875, "tRecall_90": 0.3, "token_acc": 0.9454892341237394, "epoch": 0.9541162829219811, "step": 2880}, {"loss": 0.144146728515625, "grad_norm": 0.33091646432876587, "learning_rate": 5.158921966693042e-08, "tIoU": 0.6037317752838135, "tRecall_50": 0.675, "tRecall_75": 0.4625, "tRecall_90": 0.29375, "token_acc": 0.9470362239297475, "epoch": 0.955772734802054, "step": 2885}, {"loss": 0.1508544921875, "grad_norm": 0.638655960559845, "learning_rate": 4.7817139133277126e-08, "tIoU": 0.5142153739929199, "tRecall_50": 0.60625, "tRecall_75": 0.3375, "tRecall_90": 0.2, "token_acc": 0.937192790824686, "epoch": 0.9574291866821268, "step": 2890}, {"loss": 0.160791015625, "grad_norm": 0.7477061152458191, "learning_rate": 4.418758431858572e-08, "tIoU": 0.5251534938812256, "tRecall_50": 0.575, "tRecall_75": 0.3625, "tRecall_90": 0.21875, "token_acc": 0.9371584699453552, "epoch": 0.9590856385621997, "step": 2895}, {"loss": 0.15504150390625, "grad_norm": 0.42086467146873474, "learning_rate": 4.070065968283354e-08, "tIoU": 0.571649169921875, "tRecall_50": 0.625, "tRecall_75": 0.43125, "tRecall_90": 0.29375, "token_acc": 0.9434271658923203, "epoch": 0.9607420904422727, "step": 2900}, {"loss": 0.150555419921875, "grad_norm": 0.7915459871292114, "learning_rate": 3.7356465581047105e-08, "tIoU": 0.5531152248382568, "tRecall_50": 0.6, "tRecall_75": 0.40625, "tRecall_90": 0.2875, "token_acc": 0.9445203607543045, "epoch": 0.9623985423223456, "step": 2905}, {"loss": 0.158453369140625, "grad_norm": 0.610054075717926, "learning_rate": 3.4155098260412744e-08, "tIoU": 0.5769317626953125, "tRecall_50": 0.6375, "tRecall_75": 0.45, "tRecall_90": 0.28125, "token_acc": 0.9402455661664393, "epoch": 0.9640549942024185, "step": 2910}, {"loss": 0.16090087890625, "grad_norm": 0.5148957967758179, "learning_rate": 3.1096649857508264e-08, "tIoU": 0.5256228923797608, "tRecall_50": 0.6, "tRecall_75": 0.39375, "tRecall_90": 0.225, "token_acc": 0.9383821932681867, "epoch": 0.9657114460824913, "step": 2915}, {"loss": 0.15223388671875, "grad_norm": 0.4604478180408478, "learning_rate": 2.8181208395648973e-08, "tIoU": 0.5296051502227783, "tRecall_50": 0.58125, "tRecall_75": 0.39375, "tRecall_90": 0.23125, "token_acc": 0.9399782726778925, "epoch": 0.9673678979625642, "step": 2920}, {"loss": 0.173046875, "grad_norm": 0.5151129364967346, "learning_rate": 2.5408857782357465e-08, "tIoU": 0.47686147689819336, "tRecall_50": 0.5, "tRecall_75": 0.3375, "tRecall_90": 0.2125, "token_acc": 0.9381668946648427, "epoch": 0.9690243498426371, "step": 2925}, {"loss": 0.1605712890625, "grad_norm": 0.6586578488349915, "learning_rate": 2.2779677806945567e-08, "tIoU": 0.5430468559265137, "tRecall_50": 0.59375, "tRecall_75": 0.41875, "tRecall_90": 0.25625, "token_acc": 0.9390710382513662, "epoch": 0.97068080172271, "step": 2930}, {"loss": 0.159130859375, "grad_norm": 0.752968966960907, "learning_rate": 2.0293744138219495e-08, "tIoU": 0.5184082508087158, "tRecall_50": 0.55625, "tRecall_75": 0.36875, "tRecall_90": 0.2125, "token_acc": 0.9366984993178717, "epoch": 0.9723372536027829, "step": 2935}, {"loss": 0.14085693359375, "grad_norm": 0.44599080085754395, "learning_rate": 1.7951128322302168e-08, "tIoU": 0.5751462936401367, "tRecall_50": 0.625, "tRecall_75": 0.4625, "tRecall_90": 0.33125, "token_acc": 0.9483747609942639, "epoch": 0.9739937054828557, "step": 2940}, {"loss": 0.161328125, "grad_norm": 0.6227806210517883, "learning_rate": 1.5751897780573734e-08, "tIoU": 0.5526154994964599, "tRecall_50": 0.60625, "tRecall_75": 0.40625, "tRecall_90": 0.28125, "token_acc": 0.9422343324250682, "epoch": 0.9756501573629286, "step": 2945}, {"loss": 0.15426025390625, "grad_norm": 0.5416595339775085, "learning_rate": 1.3696115807732002e-08, "tIoU": 0.5484757900238038, "tRecall_50": 0.625, "tRecall_75": 0.43125, "tRecall_90": 0.24375, "token_acc": 0.940677966101695, "epoch": 0.9773066092430015, "step": 2950}, {"loss": 0.1627685546875, "grad_norm": 1.0069711208343506, "learning_rate": 1.1783841569968368e-08, "tIoU": 0.5158636093139648, "tRecall_50": 0.575, "tRecall_75": 0.35625, "tRecall_90": 0.25625, "token_acc": 0.9338255400601586, "epoch": 0.9789630611230744, "step": 2955}, {"loss": 0.15572509765625, "grad_norm": 0.7520166039466858, "learning_rate": 1.0015130103268044e-08, "tIoU": 0.5539876461029053, "tRecall_50": 0.58125, "tRecall_75": 0.4, "tRecall_90": 0.29375, "token_acc": 0.9455391351943077, "epoch": 0.9806195130031473, "step": 2960}, {"loss": 0.1467529296875, "grad_norm": 0.47164127230644226, "learning_rate": 8.390032311824115e-09, "tIoU": 0.5772971153259278, "tRecall_50": 0.625, "tRecall_75": 0.4125, "tRecall_90": 0.30625, "token_acc": 0.9446564885496184, "epoch": 0.9822759648832201, "step": 2965}, {"loss": 0.16009521484375, "grad_norm": 0.5869075655937195, "learning_rate": 6.908594966573701e-09, "tIoU": 0.5448875427246094, "tRecall_50": 0.60625, "tRecall_75": 0.4, "tRecall_90": 0.28125, "token_acc": 0.939945652173913, "epoch": 0.983932416763293, "step": 2970}, {"loss": 0.1591552734375, "grad_norm": 0.7166591882705688, "learning_rate": 5.570860703850711e-09, "tIoU": 0.5079808712005616, "tRecall_50": 0.5375, "tRecall_75": 0.3875, "tRecall_90": 0.21875, "token_acc": 0.938113413304253, "epoch": 0.9855888686433659, "step": 2975}, {"loss": 0.14058837890625, "grad_norm": 0.4753721356391907, "learning_rate": 4.3768680241590425e-09, "tIoU": 0.5579365253448486, "tRecall_50": 0.6, "tRecall_75": 0.43125, "tRecall_90": 0.325, "token_acc": 0.946195652173913, "epoch": 0.9872453205234388, "step": 2980}, {"loss": 0.1630615234375, "grad_norm": 0.6353902220726013, "learning_rate": 3.3266512910651303e-09, "tIoU": 0.5626879692077636, "tRecall_50": 0.60625, "tRecall_75": 0.3875, "tRecall_90": 0.24375, "token_acc": 0.9439912996193583, "epoch": 0.9889017724035116, "step": 2985}, {"loss": 0.162890625, "grad_norm": 0.7356439828872681, "learning_rate": 2.4202407302093e-09, "tIoU": 0.5033055305480957, "tRecall_50": 0.5625, "tRecall_75": 0.36875, "tRecall_90": 0.20625, "token_acc": 0.9392039258451472, "epoch": 0.9905582242835845, "step": 2990}, {"loss": 0.16402587890625, "grad_norm": 0.4798955023288727, "learning_rate": 1.657662428434792e-09, "tIoU": 0.5145360469818115, "tRecall_50": 0.55625, "tRecall_75": 0.33125, "tRecall_90": 0.25, "token_acc": 0.9397820163487739, "epoch": 0.9922146761636574, "step": 2995}, {"loss": 0.1508544921875, "grad_norm": 0.8724395632743835, "learning_rate": 1.0389383330361436e-09, "tIoU": 0.5828092575073243, "tRecall_50": 0.65625, "tRecall_75": 0.4375, "tRecall_90": 0.2625, "token_acc": 0.9477693144722524, "epoch": 0.9938711280437303, "step": 3000}, {"loss": 0.161175537109375, "grad_norm": 0.4469767212867737, "learning_rate": 5.64086251130247e-10, "tIoU": 0.514407730102539, "tRecall_50": 0.5, "tRecall_75": 0.375, "tRecall_90": 0.275, "token_acc": 0.9394843962008141, "epoch": 0.9955275799238033, "step": 3005}, {"loss": 0.178851318359375, "grad_norm": 0.6539129614830017, "learning_rate": 2.331198491417608e-10, "tIoU": 0.4971726894378662, "tRecall_50": 0.5375, "tRecall_75": 0.35, "tRecall_90": 0.19375, "token_acc": 0.9323553382233088, "epoch": 0.997184031803876, "step": 3010}, {"loss": 0.175439453125, "grad_norm": 0.34182894229888916, "learning_rate": 4.604865241064626e-11, "tIoU": 0.5114048480987549, "tRecall_50": 0.54375, "tRecall_75": 0.36875, "tRecall_90": 0.26875, "token_acc": 0.9353085077466703, "epoch": 0.998840483683949, "step": 3015}, {"train_runtime": 40919.318, "train_samples_per_second": 2.36, "train_steps_per_second": 0.074, "total_flos": 8.424853533524125e+19, "train_loss": 0.1892424182569637, "tIoU": 0.5681568554469517, "tRecall_50": 0.6160714285714286, "tRecall_75": 0.42857142857142855, "tRecall_90": 0.25, "token_acc": 0.9433884297520662, "epoch": 1.0, "step": 3019}], "memory": 40.099609375} diff --git a/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/added_tokens.json b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/chat_template.jinja b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/config.json b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/config.json new file mode 100644 index 0000000000000000000000000000000000000000..adb449916091c26098a94af0fa4add48558a9494 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/config.json @@ -0,0 +1,65 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "bfloat16", + "image_token_id": 151655, + "model_type": "qwen3_vl", + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "bfloat16", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9728, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_type": "default" + }, + "rope_theta": 5000000, + "tie_word_embeddings": true, + "use_cache": true, + "vocab_size": 151936 + }, + "tie_word_embeddings": true, + "transformers_version": "4.57.3", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 5, + 11, + 17 + ], + "depth": 24, + "dtype": "bfloat16", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} diff --git a/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/generation_config.json b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..3dd2c545b96b4d51586680d7bb27fdca5c90dfe9 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "4.57.3" +} diff --git a/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/model.safetensors.index.json b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..58d118e8ce51a42ed3b147a55f9c92f39d273c60 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/model.safetensors.index.json @@ -0,0 +1,721 @@ +{ + "metadata": { + "total_parameters": 4437815808, + "total_size": 8875631616 + }, + "weight_map": { + "model.language_model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.patch_embed.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.patch_embed.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.pos_embed.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/preprocessor_config.json b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..2fa6553554f69b4e3a6b742a75264ff9dff90ad0 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/preprocessor_config.json @@ -0,0 +1,39 @@ +{ + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 +} diff --git a/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/special_tokens_map.json b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/tokenizer_config.json b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/video_preprocessor_config.json b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/video_preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..e32b1d90356f74457edbfca27aa1e121dab7cdc7 --- /dev/null +++ b/posttrain_models/qwen3vl_4b_timelens100k_epoch1_lora32_batch124578_epoch4/video_preprocessor_config.json @@ -0,0 +1,41 @@ +{ + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/added_tokens.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/chat_template.jinja b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ff29b27714d520c3f1a5e08917e084592272be3b --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/config.json @@ -0,0 +1,67 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "float32", + "eos_token_id": 151645, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "float32", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 4096, + "initializer_range": 0.02, + "intermediate_size": 12288, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_theta": 5000000, + "use_cache": true, + "vocab_size": 151936, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_theta": 5000000, + "rope_type": "default" + } + }, + "tie_word_embeddings": false, + "transformers_version": "5.0.0.dev0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 8, + 16, + 24 + ], + "depth": 27, + "dtype": "float32", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1152, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4304, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 4096, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/generation_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/model.safetensors.index.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..0e2351752d45fddcd3f8e8ba09c179a9e17674fc --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/model.safetensors.index.json @@ -0,0 +1,758 @@ +{ + "metadata": { + "total_parameters": 8767123696, + "total_size": 17534247392 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00004.safetensors", + "model.language_model.embed_tokens.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.norm.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.25.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00004-of-00004.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.merger.norm.bias": "model-00001-of-00004.safetensors", + "model.visual.merger.norm.weight": "model-00001-of-00004.safetensors", + "model.visual.patch_embed.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.patch_embed.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.pos_embed.weight": "model-00004-of-00004.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/processor_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/special_tokens_map.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/tokenizer_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_10/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/added_tokens.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/chat_template.jinja b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ff29b27714d520c3f1a5e08917e084592272be3b --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/config.json @@ -0,0 +1,67 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "float32", + "eos_token_id": 151645, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "float32", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 4096, + "initializer_range": 0.02, + "intermediate_size": 12288, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_theta": 5000000, + "use_cache": true, + "vocab_size": 151936, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_theta": 5000000, + "rope_type": "default" + } + }, + "tie_word_embeddings": false, + "transformers_version": "5.0.0.dev0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 8, + 16, + 24 + ], + "depth": 27, + "dtype": "float32", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1152, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4304, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 4096, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/generation_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/model.safetensors.index.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..9427bc7662b18bda551e4cc9b77d1c98e3f2dc3f --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/model.safetensors.index.json @@ -0,0 +1,758 @@ +{ + "metadata": { + "total_parameters": 8767123696, + "total_size": 17534247392 + }, + "weight_map": { + "lm_head.weight": "model-00004-of-00004.safetensors", + "model.language_model.embed_tokens.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.norm.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00002-of-00004.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.merger.norm.bias": "model-00001-of-00004.safetensors", + "model.visual.merger.norm.weight": "model-00002-of-00004.safetensors", + "model.visual.patch_embed.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.patch_embed.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.pos_embed.weight": "model-00002-of-00004.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/processor_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/special_tokens_map.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/tokenizer_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_100/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/added_tokens.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/chat_template.jinja b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ff29b27714d520c3f1a5e08917e084592272be3b --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/config.json @@ -0,0 +1,67 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "float32", + "eos_token_id": 151645, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "float32", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 4096, + "initializer_range": 0.02, + "intermediate_size": 12288, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_theta": 5000000, + "use_cache": true, + "vocab_size": 151936, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_theta": 5000000, + "rope_type": "default" + } + }, + "tie_word_embeddings": false, + "transformers_version": "5.0.0.dev0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 8, + 16, + 24 + ], + "depth": 27, + "dtype": "float32", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1152, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4304, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 4096, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/generation_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/model.safetensors.index.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..d8724472a6b81671a1cfde27454ffa42a95dbaf4 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/model.safetensors.index.json @@ -0,0 +1,758 @@ +{ + "metadata": { + "total_parameters": 8767123696, + "total_size": 17534247392 + }, + "weight_map": { + "lm_head.weight": "model-00004-of-00004.safetensors", + "model.language_model.embed_tokens.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.norm.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.25.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.25.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00003-of-00004.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.merger.norm.bias": "model-00001-of-00004.safetensors", + "model.visual.merger.norm.weight": "model-00003-of-00004.safetensors", + "model.visual.patch_embed.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.patch_embed.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.pos_embed.weight": "model-00002-of-00004.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/processor_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/special_tokens_map.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/tokenizer_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_101/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/added_tokens.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/chat_template.jinja b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ff29b27714d520c3f1a5e08917e084592272be3b --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/config.json @@ -0,0 +1,67 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "float32", + "eos_token_id": 151645, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "float32", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 4096, + "initializer_range": 0.02, + "intermediate_size": 12288, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_theta": 5000000, + "use_cache": true, + "vocab_size": 151936, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_theta": 5000000, + "rope_type": "default" + } + }, + "tie_word_embeddings": false, + "transformers_version": "5.0.0.dev0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 8, + 16, + 24 + ], + "depth": 27, + "dtype": "float32", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1152, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4304, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 4096, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/generation_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/model.safetensors.index.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..6e7c7ffcd1c8da3746025a6237ec866a0d5efd41 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/model.safetensors.index.json @@ -0,0 +1,758 @@ +{ + "metadata": { + "total_parameters": 8767123696, + "total_size": 17534247392 + }, + "weight_map": { + "lm_head.weight": "model-00004-of-00004.safetensors", + "model.language_model.embed_tokens.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.norm.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.25.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00001-of-00004.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.merger.norm.bias": "model-00002-of-00004.safetensors", + "model.visual.merger.norm.weight": "model-00001-of-00004.safetensors", + "model.visual.patch_embed.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.patch_embed.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.pos_embed.weight": "model-00003-of-00004.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/processor_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/special_tokens_map.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/tokenizer_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_30/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/added_tokens.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/chat_template.jinja b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ff29b27714d520c3f1a5e08917e084592272be3b --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/config.json @@ -0,0 +1,67 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "float32", + "eos_token_id": 151645, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "float32", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 4096, + "initializer_range": 0.02, + "intermediate_size": 12288, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_theta": 5000000, + "use_cache": true, + "vocab_size": 151936, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_theta": 5000000, + "rope_type": "default" + } + }, + "tie_word_embeddings": false, + "transformers_version": "5.0.0.dev0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 8, + 16, + 24 + ], + "depth": 27, + "dtype": "float32", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1152, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4304, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 4096, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/generation_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/model.safetensors.index.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..e7c3e3d4e7018abe7c4b8aeacf8757d89dc6f4e0 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/model.safetensors.index.json @@ -0,0 +1,758 @@ +{ + "metadata": { + "total_parameters": 8767123696, + "total_size": 17534247392 + }, + "weight_map": { + "lm_head.weight": "model-00004-of-00004.safetensors", + "model.language_model.embed_tokens.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.norm.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.25.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.25.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00002-of-00004.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.merger.norm.bias": "model-00004-of-00004.safetensors", + "model.visual.merger.norm.weight": "model-00001-of-00004.safetensors", + "model.visual.patch_embed.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.patch_embed.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.pos_embed.weight": "model-00003-of-00004.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/processor_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/special_tokens_map.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/tokenizer_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_50/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/added_tokens.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/chat_template.jinja b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ff29b27714d520c3f1a5e08917e084592272be3b --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/config.json @@ -0,0 +1,67 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "float32", + "eos_token_id": 151645, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "float32", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 4096, + "initializer_range": 0.02, + "intermediate_size": 12288, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_theta": 5000000, + "use_cache": true, + "vocab_size": 151936, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_theta": 5000000, + "rope_type": "default" + } + }, + "tie_word_embeddings": false, + "transformers_version": "5.0.0.dev0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 8, + 16, + 24 + ], + "depth": 27, + "dtype": "float32", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1152, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4304, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 4096, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/generation_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/model.safetensors.index.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..9e94f52d50cef4e3292257db9a0f504aff2105c0 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/model.safetensors.index.json @@ -0,0 +1,758 @@ +{ + "metadata": { + "total_parameters": 8767123696, + "total_size": 17534247392 + }, + "weight_map": { + "lm_head.weight": "model-00001-of-00004.safetensors", + "model.language_model.embed_tokens.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.norm.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00003-of-00004.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.merger.norm.bias": "model-00002-of-00004.safetensors", + "model.visual.merger.norm.weight": "model-00004-of-00004.safetensors", + "model.visual.patch_embed.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.patch_embed.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.pos_embed.weight": "model-00003-of-00004.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/processor_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/special_tokens_map.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/tokenizer_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_70/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/added_tokens.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/chat_template.jinja b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ff29b27714d520c3f1a5e08917e084592272be3b --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/config.json @@ -0,0 +1,67 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "float32", + "eos_token_id": 151645, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "float32", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 4096, + "initializer_range": 0.02, + "intermediate_size": 12288, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "rms_norm_eps": 1e-06, + "rope_theta": 5000000, + "use_cache": true, + "vocab_size": 151936, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_theta": 5000000, + "rope_type": "default" + } + }, + "tie_word_embeddings": false, + "transformers_version": "5.0.0.dev0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 8, + 16, + 24 + ], + "depth": 27, + "dtype": "float32", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1152, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4304, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 4096, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} \ No newline at end of file diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/generation_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..8458f29aec2209e3f63ca59bca95814310a80de1 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.0.0.dev0" +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/model.safetensors.index.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..d8096f9956ed18143abdee20bb68e564d666bd5e --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/model.safetensors.index.json @@ -0,0 +1,758 @@ +{ + "metadata": { + "total_parameters": 8767123696, + "total_size": 17534247392 + }, + "weight_map": { + "lm_head.weight": "model-00003-of-00004.safetensors", + "model.language_model.embed_tokens.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00002-of-00004.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00003-of-00004.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.language_model.norm.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.24.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.24.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.24.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.attn.proj.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.attn.qkv.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.25.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.25.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.attn.proj.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.26.mlp.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.26.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.26.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.26.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00002-of-00004.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00004-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00001-of-00004.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00003-of-00004.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00004-of-00004.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00001-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00003-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00004-of-00004.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00001-of-00004.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00003-of-00004.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00003-of-00004.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00004-of-00004.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00002-of-00004.safetensors", + "model.visual.merger.norm.bias": "model-00001-of-00004.safetensors", + "model.visual.merger.norm.weight": "model-00003-of-00004.safetensors", + "model.visual.patch_embed.proj.bias": "model-00002-of-00004.safetensors", + "model.visual.patch_embed.proj.weight": "model-00001-of-00004.safetensors", + "model.visual.pos_embed.weight": "model-00003-of-00004.safetensors" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/processor_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a4fe483a685ede297970ea7a13145d5aabe63a1c --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/processor_config.json @@ -0,0 +1,85 @@ +{ + "image_processor": { + "crop_size": null, + "data_format": "channels_first", + "device": null, + "disable_grouping": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_seq_length": null, + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_pixels": null, + "merge_size": 2, + "min_pixels": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "return_tensors": null, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/special_tokens_map.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/tokenizer_config.json b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/qwen3vl_8b_ntr_multimath_gemini_50k/global_step_90/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/replan_qwen2_5_vl_7b/.gitattributes b/posttrain_models/replan_qwen2_5_vl_7b/.gitattributes new file mode 100644 index 0000000000000000000000000000000000000000..ed4d03266b88b246c70eb815d4b673b2bc614c58 --- /dev/null +++ b/posttrain_models/replan_qwen2_5_vl_7b/.gitattributes @@ -0,0 +1,39 @@ +*.7z filter=lfs diff=lfs merge=lfs -text +*.arrow filter=lfs diff=lfs merge=lfs -text +*.bin filter=lfs diff=lfs merge=lfs -text +*.bz2 filter=lfs diff=lfs merge=lfs -text +*.ckpt filter=lfs diff=lfs merge=lfs -text +*.ftz filter=lfs diff=lfs merge=lfs -text +*.gz filter=lfs diff=lfs merge=lfs -text +*.h5 filter=lfs diff=lfs merge=lfs -text +*.joblib filter=lfs diff=lfs merge=lfs -text +*.lfs.* filter=lfs diff=lfs merge=lfs -text +*.mlmodel filter=lfs diff=lfs merge=lfs -text +*.model filter=lfs diff=lfs merge=lfs -text +*.msgpack filter=lfs diff=lfs merge=lfs -text +*.npy filter=lfs diff=lfs merge=lfs -text +*.npz filter=lfs diff=lfs merge=lfs -text +*.onnx filter=lfs diff=lfs merge=lfs -text +*.ot filter=lfs diff=lfs merge=lfs -text +*.parquet filter=lfs diff=lfs merge=lfs -text +*.pb filter=lfs diff=lfs merge=lfs -text +*.pickle filter=lfs diff=lfs merge=lfs -text +*.pkl filter=lfs diff=lfs merge=lfs -text +*.pt filter=lfs diff=lfs merge=lfs -text +*.pth filter=lfs diff=lfs merge=lfs -text +*.rar filter=lfs diff=lfs merge=lfs -text +*.safetensors filter=lfs diff=lfs merge=lfs -text +saved_model/**/* filter=lfs diff=lfs merge=lfs -text +*.tar.* filter=lfs diff=lfs merge=lfs -text +*.tar filter=lfs diff=lfs merge=lfs -text +*.tflite filter=lfs diff=lfs merge=lfs -text +*.tgz filter=lfs diff=lfs merge=lfs -text +*.wasm filter=lfs diff=lfs merge=lfs -text +*.xz filter=lfs diff=lfs merge=lfs -text +*.zip filter=lfs diff=lfs merge=lfs -text +*.zst filter=lfs diff=lfs merge=lfs -text +*tfevents* filter=lfs diff=lfs merge=lfs -text +sft(1).jsonl filter=lfs diff=lfs merge=lfs -text +tokenizer.json filter=lfs diff=lfs merge=lfs -text +assets/cup.png filter=lfs diff=lfs merge=lfs -text +assets/teaser.jpg filter=lfs diff=lfs merge=lfs -text diff --git a/posttrain_models/replan_qwen2_5_vl_7b/README.md b/posttrain_models/replan_qwen2_5_vl_7b/README.md new file mode 100644 index 0000000000000000000000000000000000000000..337bd1dac26c673d76c5b5877fa7a55b3df218e4 --- /dev/null +++ b/posttrain_models/replan_qwen2_5_vl_7b/README.md @@ -0,0 +1,171 @@ +--- +base_model: +- Qwen/Qwen2.5-VL-7B-Instruct +pipeline_tag: image-to-text +library_name: transformers +license: apache-2.0 +--- + +# REPLAN: REASONING-GUIDED REGION PLANNING FOR COMPLEX INSTRUCTION-BASED IMAGE EDITING + +This model is part of the work presented in the paper [RePlan: Reasoning-guided Region Planning for Complex Instruction-based Image Editing](https://huggingface.co/papers/2512.16864). + +

+ + Project Page +      + + Paper (Hugging Face) +      + + Paper (arXiv) +      + + GitHub Code +      + + Hugging Face Dataset +      + + Hugging Face Model + +

+ + + +Demo video of RePlan: +
+ +
+ + +## Model Summary + +This model is the **Planner module** of the **RePlan** framework, designed for complex instruction-based image editing. It is a fine-tuned version of **Qwen2.5-VL-7B**, trained with GRPO and only ~1k samples without any paired images. + +Given an input image and a natural language editing instruction, this model performs Chain-of-Thought (CoT) reasoning to decompose the task. It outputs structured guidance containing: + +1. **Reasoning:** Analysis of the image and instruction. +2. **Global Edits:** Instructions for the entire image (if necessary). +3. **Regional Edits:** Precise bounding boxes (`bbox_2d`) and specific prompts (`hint`) for local regions. + + +## Paper Introduction + +**Paper Title:** RePlan: Reasoning-Guided Region Planning for Complex Instruction-Based Image Editing + +Existing instruction-based image editing models often struggle with **Instruction-Visual Complexity (IV-Complexity)**: scenarios involving cluttered visuals, ambiguous instructions, or the need for multi-step reasoning. + +**RePlan** introduces a "Plan-then-Execute" strategy: + +1. **Plan:** This VLM planner analyzes the scene, grounds the instruction to specific pixels, and generates a precise editing plan. +2. **Execute:** A diffusion model (equipped with a Training-Free Attention Region Injection mechanism) applies the edits based on the planner's guidance. + +Experiments show that RePlan significantly outperforms baselines in visual reasoning and background consistency. + + +## Usage + +### System Prompt + +**Crucial:** To get the correct XML/JSON structured output, you **MUST** use the following System Prompt. + +```text +You are an expert AI image editing assistant. Your task is to carefully analyze a user's editing instruction and input image, reason step by step, and then decompose the necessary actions into global and local edits. + +### Rules +1. **Global Edits:** Affect the entire image's style, lighting, color grading, or overall composition. Global edits should only be derived if they are essential for achieving the user's core instruction. +2. **Local Edits:** Target specific objects or areas. These instructions go into the `` tag. +3. **Hint Quality:** The `hint` text MUST be a concise, visually descriptive instruction for its specific region. It should clearly state the expected visual outcome. +4. **Strict Separation:** Instructions for local edits in `` MUST NOT be duplicated in the `` prompt. +5. **Edge Case - No Global Edits:** If no global edits are necessary to achieve the user's goal, `` MUST be placeholder ‘keep remaining part of image unchanged.’. +6. **Edge Case - No Local Edits:** If no local edits are needed, `` must be an empty list `[]`. + +### Output Format +Your entire output must follow this format, with no text outside the tags. + +Reasoning processGlobal edit instruction[{"bbox_2d": [10,150,150,210], "point_2d": [30,175], "hint": "change the color of this one apple to blue"}, {"bbox_2d": [150,50,200,150], "point_2d": [175,75], "hint": "keep this one apple unchanged"}] + +``` + +### Inference Code Example + +```python +import torch +from transformers import AutoProcessor, Qwen2_5_VLForConditionalGeneration +from qwen_vl_utils import process_vision_info + +# 1. Load Model +model_path = "path/to/your/model" # Replace with huggingface repo id +model = Qwen2_5_VLForConditionalGeneration.from_pretrained( + model_path, torch_dtype=torch.float16, device_map="auto" +) +processor = AutoProcessor.from_pretrained(model_path) + +# 2. Define System Prompt (See section above) +SYSTEM_PROMPT = "..." # Paste the full system prompt here + +# 3. Prepare Inputs +image_path = "./example_image.jpg" +user_instruction = "Replace the cup that has been used and left on the desk with a small potted plant." + +messages = [ + {"role": "system", "content": SYSTEM_PROMPT}, + { + "role": "user", + "content": [ + {"type": "image", "image": image_path}, + {"type": "text", "text": user_instruction}, + ], + }, +] + +# 4. Processing +text = processor.apply_chat_template( + messages, tokenize=False, add_generation_prompt=True +) + +image_inputs, video_inputs = process_vision_info(messages) +inputs = processor( + text=[text], + images=image_inputs, + videos=video_inputs, + padding=True, + return_tensors="pt", +).to(model.device) + +# 5. Generate +generated_ids = model.generate(**inputs, max_new_tokens=2048) +generated_ids_trimmed = [ + out_ids[len(in_ids):] for in_ids, out_ids in zip(inputs.input_ids, generated_ids) +] +output_text_list = processor.batch_decode( + generated_ids_trimmed, + skip_special_tokens=True, + clean_up_tokenization_spaces=False +) + +# Output is the raw XML string containing , , and +response_text = output_text_list[0] if isinstance(output_text_list, list) else output_text_list +print(response_text) + +``` + +## Input / Output Examples + +### Example 1: Complex Reasoning + +**Input Image:** + + + +**Instruction:** "Replace the cup that has been used and left on the desk with a small potted plant." + +**Output Format Example:** + +```xml + +The user wants to replace a "used" cup. I see two vessels: a red coffee cup and a glass of water. The red cup appears empty and discarded, fitting the "used and left" description. The glass of water is still potentially in use. Therefore, I should target the red cup for replacement and explicitly keep the glass unchanged to avoid confusion. +keep remaining part of image unchanged.[{"bbox_2d": [224, 372, 263, 431], "hint": "Replace this red cup with a small potted plant"}, {"bbox_2d": [175, 329, 220, 388], "hint": "Keep this glass unchanged"}] + +``` \ No newline at end of file diff --git a/posttrain_models/replan_qwen2_5_vl_7b/added_tokens.json b/posttrain_models/replan_qwen2_5_vl_7b/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..482ced4679301bf287ebb310bdd1790eb4514232 --- /dev/null +++ b/posttrain_models/replan_qwen2_5_vl_7b/added_tokens.json @@ -0,0 +1,24 @@ +{ + "
": 151658, + "": 151657, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/replan_qwen2_5_vl_7b/chat_template.jinja b/posttrain_models/replan_qwen2_5_vl_7b/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..6c226632394ae7474b0d4b13e15793eac2e21ee9 --- /dev/null +++ b/posttrain_models/replan_qwen2_5_vl_7b/chat_template.jinja @@ -0,0 +1,7 @@ +{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system +You are a helpful assistant.<|im_end|> +{% endif %}<|im_start|>{{ message['role'] }} +{% if message['content'] is string %}{{ message['content'] }}<|im_end|> +{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|> +{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant +{% endif %} \ No newline at end of file diff --git a/posttrain_models/replan_qwen2_5_vl_7b/config.json b/posttrain_models/replan_qwen2_5_vl_7b/config.json new file mode 100644 index 0000000000000000000000000000000000000000..94699ded96c28ff66c52cfcbe3559c1fdf68fa2d --- /dev/null +++ b/posttrain_models/replan_qwen2_5_vl_7b/config.json @@ -0,0 +1,105 @@ +{ + "architectures": [ + "Qwen2_5_VLForConditionalGeneration" + ], + "attention_dropout": 0.0, + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 3584, + "image_token_id": 151655, + "initializer_range": 0.02, + "intermediate_size": 18944, + "max_position_embeddings": 128000, + "max_window_layers": 28, + "model_type": "qwen2_5_vl", + "num_attention_heads": 28, + "num_hidden_layers": 28, + "num_key_value_heads": 4, + "pad_token_id": 151643, + "rms_norm_eps": 1e-06, + "rope_scaling": { + "mrope_section": [ + 16, + 24, + 24 + ], + "rope_type": "default", + "type": "default" + }, + "rope_theta": 1000000.0, + "sliding_window": 32768, + "text_config": { + "architectures": [ + "Qwen2_5_VLForConditionalGeneration" + ], + "attention_dropout": 0.0, + "bos_token_id": 151643, + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 3584, + "image_token_id": null, + "initializer_range": 0.02, + "intermediate_size": 18944, + "max_position_embeddings": 128000, + "max_window_layers": 28, + "model_type": "qwen2_5_vl_text", + "num_attention_heads": 28, + "num_hidden_layers": 28, + "num_key_value_heads": 4, + "rms_norm_eps": 1e-06, + "rope_scaling": { + "mrope_section": [ + 16, + 24, + 24 + ], + "rope_type": "default", + "type": "default" + }, + "rope_theta": 1000000.0, + "sliding_window": 32768, + "torch_dtype": "float32", + "use_cache": true, + "use_sliding_window": false, + "video_token_id": null, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652, + "vision_token_id": 151654, + "vocab_size": 152064 + }, + "tie_word_embeddings": false, + "torch_dtype": "float32", + "transformers_version": "4.52.4", + "use_cache": true, + "use_sliding_window": false, + "video_token_id": 151656, + "vision_config": { + "depth": 32, + "fullatt_block_indexes": [ + 7, + 15, + 23, + 31 + ], + "hidden_act": "silu", + "hidden_size": 1280, + "in_channels": 3, + "in_chans": 3, + "initializer_range": 0.02, + "intermediate_size": 3420, + "model_type": "qwen2_5_vl", + "num_heads": 16, + "out_hidden_size": 3584, + "patch_size": 14, + "spatial_merge_size": 2, + "spatial_patch_size": 14, + "temporal_patch_size": 2, + "tokens_per_second": 2, + "torch_dtype": "float32", + "window_size": 112 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652, + "vision_token_id": 151654, + "vocab_size": 152064 +} diff --git a/posttrain_models/replan_qwen2_5_vl_7b/generation_config.json b/posttrain_models/replan_qwen2_5_vl_7b/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0b1c4ed064c2bc213e08c49a93de5f0a403b7b7a --- /dev/null +++ b/posttrain_models/replan_qwen2_5_vl_7b/generation_config.json @@ -0,0 +1,7 @@ +{ + "_from_model_config": true, + "bos_token_id": 151643, + "eos_token_id": 151645, + "pad_token_id": 151643, + "transformers_version": "4.52.4" +} diff --git a/posttrain_models/replan_qwen2_5_vl_7b/model.safetensors.index.json b/posttrain_models/replan_qwen2_5_vl_7b/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..061e9b975965002aafb1f581af60dbaaec53e60c --- /dev/null +++ b/posttrain_models/replan_qwen2_5_vl_7b/model.safetensors.index.json @@ -0,0 +1,736 @@ +{ + "metadata": { + "total_size": 16584333312 + }, + "weight_map": { + "lm_head.weight": "model-00001-of-00004.safetensors", + "model.embed_tokens.weight": "model-00003-of-00004.safetensors", + "model.layers.0.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.0.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.0.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.0.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.0.self_attn.v_proj.bias": "model-00004-of-00004.safetensors", + "model.layers.0.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.1.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.1.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.1.self_attn.v_proj.bias": "model-00004-of-00004.safetensors", + "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.10.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.10.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.10.self_attn.k_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.10.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.10.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.10.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.10.self_attn.v_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.10.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.11.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.11.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.11.self_attn.q_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.11.self_attn.v_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.11.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.12.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.12.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.12.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.12.self_attn.k_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.12.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.12.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.12.self_attn.v_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.12.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.13.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.13.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.13.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.layers.13.self_attn.k_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.13.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.13.self_attn.q_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.13.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.13.self_attn.v_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.13.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.14.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.14.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.14.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.14.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.14.self_attn.q_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.14.self_attn.v_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.14.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.15.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.15.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.15.self_attn.k_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.15.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.15.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.15.self_attn.v_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.15.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.16.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.16.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.16.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.16.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.16.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.16.self_attn.v_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.16.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.17.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.17.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.17.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.17.self_attn.k_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.17.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.17.self_attn.q_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.17.self_attn.v_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.17.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.18.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.18.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.18.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.18.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.18.self_attn.v_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.19.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.19.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.19.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.19.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.19.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.19.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.19.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.19.self_attn.v_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.19.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.2.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.2.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.2.self_attn.q_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.2.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.2.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.20.input_layernorm.weight": "model-00004-of-00004.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.20.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.20.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.20.self_attn.k_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.20.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.20.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.20.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.20.self_attn.v_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.20.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.21.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.21.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.21.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.21.self_attn.k_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.21.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.21.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.21.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.21.self_attn.v_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.21.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.22.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.22.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.layers.22.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.22.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.22.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.22.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.22.self_attn.v_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.23.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.23.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.23.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.23.self_attn.q_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.23.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.23.self_attn.v_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.23.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.24.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.24.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.24.self_attn.k_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.24.self_attn.k_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.24.self_attn.q_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.24.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.24.self_attn.v_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.24.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.25.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.25.self_attn.k_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.25.self_attn.q_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.25.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.25.self_attn.v_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.25.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.26.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.26.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.26.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.26.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.26.self_attn.q_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.26.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.26.self_attn.v_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.26.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.27.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.27.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.27.self_attn.q_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.27.self_attn.q_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.27.self_attn.v_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.3.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.3.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.3.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.3.self_attn.k_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.3.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.3.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.3.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.3.self_attn.v_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.4.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.4.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.4.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00004-of-00004.safetensors", + "model.layers.4.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.4.self_attn.k_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.4.self_attn.q_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.4.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.4.self_attn.v_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.5.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.5.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.5.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.5.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.5.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.5.self_attn.q_proj.bias": "model-00004-of-00004.safetensors", + "model.layers.5.self_attn.q_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.5.self_attn.v_proj.bias": "model-00001-of-00004.safetensors", + "model.layers.5.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.6.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.6.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.6.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.6.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.6.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.6.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.6.self_attn.q_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.6.self_attn.v_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.6.self_attn.v_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00004.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.7.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.7.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.7.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.7.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.7.self_attn.q_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.7.self_attn.v_proj.bias": "model-00004-of-00004.safetensors", + "model.layers.7.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.8.input_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.8.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.8.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.8.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.8.self_attn.k_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.8.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.8.self_attn.v_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.8.self_attn.v_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.9.input_layernorm.weight": "model-00002-of-00004.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.9.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.9.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00003-of-00004.safetensors", + "model.layers.9.self_attn.k_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.9.self_attn.k_proj.weight": "model-00002-of-00004.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00004-of-00004.safetensors", + "model.layers.9.self_attn.q_proj.bias": "model-00002-of-00004.safetensors", + "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00004.safetensors", + "model.layers.9.self_attn.v_proj.bias": "model-00003-of-00004.safetensors", + "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00004.safetensors", + "model.norm.weight": "model-00002-of-00004.safetensors", + "visual.blocks.0.attn.proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.0.attn.proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.0.attn.qkv.bias": "model-00001-of-00004.safetensors", + "visual.blocks.0.attn.qkv.weight": "model-00003-of-00004.safetensors", + "visual.blocks.0.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.0.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.0.mlp.gate_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.0.mlp.up_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.0.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.0.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.0.norm2.weight": "model-00003-of-00004.safetensors", + "visual.blocks.1.attn.proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.1.attn.proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.1.attn.qkv.bias": "model-00003-of-00004.safetensors", + "visual.blocks.1.attn.qkv.weight": "model-00003-of-00004.safetensors", + "visual.blocks.1.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.1.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.1.mlp.gate_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.1.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.1.mlp.up_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.1.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.1.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.1.norm2.weight": "model-00002-of-00004.safetensors", + "visual.blocks.10.attn.proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.10.attn.proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.10.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.10.attn.qkv.weight": "model-00002-of-00004.safetensors", + "visual.blocks.10.mlp.down_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.10.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.10.mlp.gate_proj.bias": "model-00004-of-00004.safetensors", + "visual.blocks.10.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.10.mlp.up_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.10.norm1.weight": "model-00002-of-00004.safetensors", + "visual.blocks.10.norm2.weight": "model-00002-of-00004.safetensors", + "visual.blocks.11.attn.proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.11.attn.proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.11.attn.qkv.bias": "model-00001-of-00004.safetensors", + "visual.blocks.11.attn.qkv.weight": "model-00001-of-00004.safetensors", + "visual.blocks.11.mlp.down_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.11.mlp.gate_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.11.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.11.mlp.up_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.11.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.11.norm1.weight": "model-00004-of-00004.safetensors", + "visual.blocks.11.norm2.weight": "model-00002-of-00004.safetensors", + "visual.blocks.12.attn.proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.12.attn.proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.12.attn.qkv.bias": "model-00001-of-00004.safetensors", + "visual.blocks.12.attn.qkv.weight": "model-00002-of-00004.safetensors", + "visual.blocks.12.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.12.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.12.mlp.gate_proj.bias": "model-00004-of-00004.safetensors", + "visual.blocks.12.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.12.mlp.up_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.12.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.12.norm1.weight": "model-00002-of-00004.safetensors", + "visual.blocks.12.norm2.weight": "model-00003-of-00004.safetensors", + "visual.blocks.13.attn.proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.13.attn.proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.13.attn.qkv.bias": "model-00003-of-00004.safetensors", + "visual.blocks.13.attn.qkv.weight": "model-00003-of-00004.safetensors", + "visual.blocks.13.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.13.mlp.gate_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.13.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.13.mlp.up_proj.bias": "model-00004-of-00004.safetensors", + "visual.blocks.13.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.13.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.13.norm2.weight": "model-00003-of-00004.safetensors", + "visual.blocks.14.attn.proj.bias": "model-00004-of-00004.safetensors", + "visual.blocks.14.attn.proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.14.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.14.attn.qkv.weight": "model-00001-of-00004.safetensors", + "visual.blocks.14.mlp.down_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.14.mlp.gate_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.14.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.14.mlp.up_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.14.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.14.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.14.norm2.weight": "model-00001-of-00004.safetensors", + "visual.blocks.15.attn.proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.15.attn.proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.15.attn.qkv.bias": "model-00001-of-00004.safetensors", + "visual.blocks.15.attn.qkv.weight": "model-00003-of-00004.safetensors", + "visual.blocks.15.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.15.mlp.gate_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.15.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.15.mlp.up_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.15.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "visual.blocks.15.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.15.norm2.weight": "model-00002-of-00004.safetensors", + "visual.blocks.16.attn.proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.16.attn.proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.16.attn.qkv.bias": "model-00004-of-00004.safetensors", + "visual.blocks.16.attn.qkv.weight": "model-00001-of-00004.safetensors", + "visual.blocks.16.mlp.down_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.16.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.16.mlp.gate_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.16.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.16.mlp.up_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.16.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.16.norm2.weight": "model-00003-of-00004.safetensors", + "visual.blocks.17.attn.proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.17.attn.proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.17.attn.qkv.bias": "model-00001-of-00004.safetensors", + "visual.blocks.17.attn.qkv.weight": "model-00003-of-00004.safetensors", + "visual.blocks.17.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.17.mlp.gate_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.17.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.17.mlp.up_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.17.norm1.weight": "model-00004-of-00004.safetensors", + "visual.blocks.17.norm2.weight": "model-00003-of-00004.safetensors", + "visual.blocks.18.attn.proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.18.attn.proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.18.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.18.attn.qkv.weight": "model-00002-of-00004.safetensors", + "visual.blocks.18.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.18.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.18.mlp.gate_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.18.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.18.mlp.up_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.18.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.18.norm1.weight": "model-00002-of-00004.safetensors", + "visual.blocks.18.norm2.weight": "model-00002-of-00004.safetensors", + "visual.blocks.19.attn.proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.19.attn.proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.19.attn.qkv.bias": "model-00003-of-00004.safetensors", + "visual.blocks.19.attn.qkv.weight": "model-00002-of-00004.safetensors", + "visual.blocks.19.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.19.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.19.mlp.gate_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.19.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.19.mlp.up_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.19.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.19.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.19.norm2.weight": "model-00003-of-00004.safetensors", + "visual.blocks.2.attn.proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.2.attn.proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.2.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.2.attn.qkv.weight": "model-00002-of-00004.safetensors", + "visual.blocks.2.mlp.down_proj.bias": "model-00004-of-00004.safetensors", + "visual.blocks.2.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.2.mlp.gate_proj.bias": "model-00004-of-00004.safetensors", + "visual.blocks.2.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.2.mlp.up_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.2.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.2.norm2.weight": "model-00003-of-00004.safetensors", + "visual.blocks.20.attn.proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.20.attn.proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.20.attn.qkv.bias": "model-00004-of-00004.safetensors", + "visual.blocks.20.attn.qkv.weight": "model-00002-of-00004.safetensors", + "visual.blocks.20.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.20.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.20.mlp.gate_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.20.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.20.mlp.up_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.20.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.20.norm1.weight": "model-00002-of-00004.safetensors", + "visual.blocks.20.norm2.weight": "model-00002-of-00004.safetensors", + "visual.blocks.21.attn.proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.21.attn.proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.21.attn.qkv.bias": "model-00003-of-00004.safetensors", + "visual.blocks.21.attn.qkv.weight": "model-00004-of-00004.safetensors", + "visual.blocks.21.mlp.down_proj.bias": "model-00004-of-00004.safetensors", + "visual.blocks.21.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "visual.blocks.21.mlp.gate_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.21.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.21.mlp.up_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.21.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.21.norm1.weight": "model-00002-of-00004.safetensors", + "visual.blocks.21.norm2.weight": "model-00004-of-00004.safetensors", + "visual.blocks.22.attn.proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.22.attn.proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.22.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.22.attn.qkv.weight": "model-00001-of-00004.safetensors", + "visual.blocks.22.mlp.down_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.22.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.22.mlp.gate_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.22.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.22.mlp.up_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.22.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "visual.blocks.22.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.22.norm2.weight": "model-00001-of-00004.safetensors", + "visual.blocks.23.attn.proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.23.attn.proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.23.attn.qkv.bias": "model-00003-of-00004.safetensors", + "visual.blocks.23.attn.qkv.weight": "model-00001-of-00004.safetensors", + "visual.blocks.23.mlp.down_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.23.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.23.mlp.gate_proj.bias": "model-00004-of-00004.safetensors", + "visual.blocks.23.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.23.mlp.up_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.23.norm1.weight": "model-00004-of-00004.safetensors", + "visual.blocks.23.norm2.weight": "model-00002-of-00004.safetensors", + "visual.blocks.24.attn.proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.24.attn.proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.24.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.24.attn.qkv.weight": "model-00003-of-00004.safetensors", + "visual.blocks.24.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.24.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.24.mlp.gate_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.24.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.24.mlp.up_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.24.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.24.norm1.weight": "model-00001-of-00004.safetensors", + "visual.blocks.24.norm2.weight": "model-00002-of-00004.safetensors", + "visual.blocks.25.attn.proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.25.attn.proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.25.attn.qkv.bias": "model-00001-of-00004.safetensors", + "visual.blocks.25.attn.qkv.weight": "model-00002-of-00004.safetensors", + "visual.blocks.25.mlp.down_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.25.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.25.mlp.gate_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.25.mlp.up_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.25.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.25.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.25.norm2.weight": "model-00003-of-00004.safetensors", + "visual.blocks.26.attn.proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.26.attn.proj.weight": "model-00004-of-00004.safetensors", + "visual.blocks.26.attn.qkv.bias": "model-00001-of-00004.safetensors", + "visual.blocks.26.attn.qkv.weight": "model-00001-of-00004.safetensors", + "visual.blocks.26.mlp.down_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.26.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.26.mlp.gate_proj.bias": "model-00004-of-00004.safetensors", + "visual.blocks.26.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.26.mlp.up_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.26.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.26.norm1.weight": "model-00002-of-00004.safetensors", + "visual.blocks.26.norm2.weight": "model-00003-of-00004.safetensors", + "visual.blocks.27.attn.proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.27.attn.proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.27.attn.qkv.bias": "model-00003-of-00004.safetensors", + "visual.blocks.27.attn.qkv.weight": "model-00003-of-00004.safetensors", + "visual.blocks.27.mlp.down_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.27.mlp.gate_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.27.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.27.mlp.up_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.27.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.27.norm1.weight": "model-00002-of-00004.safetensors", + "visual.blocks.27.norm2.weight": "model-00002-of-00004.safetensors", + "visual.blocks.28.attn.proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.28.attn.proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.28.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.28.attn.qkv.weight": "model-00003-of-00004.safetensors", + "visual.blocks.28.mlp.down_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.28.mlp.down_proj.weight": "model-00004-of-00004.safetensors", + "visual.blocks.28.mlp.gate_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.28.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.28.mlp.up_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.28.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.28.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.28.norm2.weight": "model-00003-of-00004.safetensors", + "visual.blocks.29.attn.proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.29.attn.proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.29.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.29.attn.qkv.weight": "model-00001-of-00004.safetensors", + "visual.blocks.29.mlp.down_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.29.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.29.mlp.gate_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.29.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.29.mlp.up_proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.29.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.29.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.29.norm2.weight": "model-00002-of-00004.safetensors", + "visual.blocks.3.attn.proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.3.attn.proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.3.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.3.attn.qkv.weight": "model-00002-of-00004.safetensors", + "visual.blocks.3.mlp.down_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.3.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.3.mlp.gate_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.3.mlp.up_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.3.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.3.norm1.weight": "model-00002-of-00004.safetensors", + "visual.blocks.3.norm2.weight": "model-00001-of-00004.safetensors", + "visual.blocks.30.attn.proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.30.attn.proj.weight": "model-00004-of-00004.safetensors", + "visual.blocks.30.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.30.attn.qkv.weight": "model-00003-of-00004.safetensors", + "visual.blocks.30.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.30.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.30.mlp.gate_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.30.mlp.gate_proj.weight": "model-00004-of-00004.safetensors", + "visual.blocks.30.mlp.up_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.30.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.30.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.30.norm2.weight": "model-00004-of-00004.safetensors", + "visual.blocks.31.attn.proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.31.attn.proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.31.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.31.attn.qkv.weight": "model-00003-of-00004.safetensors", + "visual.blocks.31.mlp.down_proj.bias": "model-00004-of-00004.safetensors", + "visual.blocks.31.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.31.mlp.gate_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.31.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.31.mlp.up_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.31.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.31.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.31.norm2.weight": "model-00001-of-00004.safetensors", + "visual.blocks.4.attn.proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.4.attn.proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.4.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.4.attn.qkv.weight": "model-00002-of-00004.safetensors", + "visual.blocks.4.mlp.down_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.4.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.4.mlp.gate_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.4.mlp.gate_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.4.mlp.up_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.4.mlp.up_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.4.norm1.weight": "model-00002-of-00004.safetensors", + "visual.blocks.4.norm2.weight": "model-00002-of-00004.safetensors", + "visual.blocks.5.attn.proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.5.attn.proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.5.attn.qkv.bias": "model-00003-of-00004.safetensors", + "visual.blocks.5.attn.qkv.weight": "model-00001-of-00004.safetensors", + "visual.blocks.5.mlp.down_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.5.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.5.mlp.gate_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.5.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.5.mlp.up_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.5.mlp.up_proj.weight": "model-00004-of-00004.safetensors", + "visual.blocks.5.norm1.weight": "model-00004-of-00004.safetensors", + "visual.blocks.5.norm2.weight": "model-00001-of-00004.safetensors", + "visual.blocks.6.attn.proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.6.attn.proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.6.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.6.attn.qkv.weight": "model-00001-of-00004.safetensors", + "visual.blocks.6.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.6.mlp.down_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.6.mlp.gate_proj.bias": "model-00003-of-00004.safetensors", + "visual.blocks.6.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.6.mlp.up_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.6.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.6.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.6.norm2.weight": "model-00001-of-00004.safetensors", + "visual.blocks.7.attn.proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.7.attn.proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.7.attn.qkv.bias": "model-00003-of-00004.safetensors", + "visual.blocks.7.attn.qkv.weight": "model-00004-of-00004.safetensors", + "visual.blocks.7.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.7.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.7.mlp.gate_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.7.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.7.mlp.up_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.7.mlp.up_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.7.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.7.norm2.weight": "model-00001-of-00004.safetensors", + "visual.blocks.8.attn.proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.8.attn.proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.8.attn.qkv.bias": "model-00003-of-00004.safetensors", + "visual.blocks.8.attn.qkv.weight": "model-00002-of-00004.safetensors", + "visual.blocks.8.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.8.mlp.down_proj.weight": "model-00002-of-00004.safetensors", + "visual.blocks.8.mlp.gate_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.8.mlp.gate_proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.8.mlp.up_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.8.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.8.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.8.norm2.weight": "model-00003-of-00004.safetensors", + "visual.blocks.9.attn.proj.bias": "model-00001-of-00004.safetensors", + "visual.blocks.9.attn.proj.weight": "model-00003-of-00004.safetensors", + "visual.blocks.9.attn.qkv.bias": "model-00002-of-00004.safetensors", + "visual.blocks.9.attn.qkv.weight": "model-00001-of-00004.safetensors", + "visual.blocks.9.mlp.down_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.9.mlp.down_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.9.mlp.gate_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.9.mlp.gate_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.9.mlp.up_proj.bias": "model-00002-of-00004.safetensors", + "visual.blocks.9.mlp.up_proj.weight": "model-00001-of-00004.safetensors", + "visual.blocks.9.norm1.weight": "model-00003-of-00004.safetensors", + "visual.blocks.9.norm2.weight": "model-00002-of-00004.safetensors", + "visual.merger.ln_q.weight": "model-00003-of-00004.safetensors", + "visual.merger.mlp.0.bias": "model-00004-of-00004.safetensors", + "visual.merger.mlp.0.weight": "model-00002-of-00004.safetensors", + "visual.merger.mlp.2.bias": "model-00001-of-00004.safetensors", + "visual.merger.mlp.2.weight": "model-00003-of-00004.safetensors", + "visual.patch_embed.proj.weight": "model-00003-of-00004.safetensors" + } +} diff --git a/posttrain_models/replan_qwen2_5_vl_7b/preprocessor_config.json b/posttrain_models/replan_qwen2_5_vl_7b/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1c234b7d9fe675bdbcff3fa77c6844ab8341e6bf --- /dev/null +++ b/posttrain_models/replan_qwen2_5_vl_7b/preprocessor_config.json @@ -0,0 +1,36 @@ +{ + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.48145466, + 0.4578275, + 0.40821073 + ], + "image_processor_type": "Qwen2VLImageProcessorFast", + "image_std": [ + 0.26862954, + 0.26130258, + 0.27577711 + ], + "input_data_format": null, + "max_pixels": 12845056, + "merge_size": 2, + "min_pixels": 3136, + "patch_size": 14, + "processor_class": "Qwen2_5_VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_tensors": null, + "size": { + "longest_edge": 12845056, + "shortest_edge": 3136 + }, + "temporal_patch_size": 2 +} diff --git a/posttrain_models/replan_qwen2_5_vl_7b/special_tokens_map.json b/posttrain_models/replan_qwen2_5_vl_7b/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/replan_qwen2_5_vl_7b/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/replan_qwen2_5_vl_7b/tokenizer_config.json b/posttrain_models/replan_qwen2_5_vl_7b/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..d595495d9de4be70a7c97263eabe3653c6492ab3 --- /dev/null +++ b/posttrain_models/replan_qwen2_5_vl_7b/tokenizer_config.json @@ -0,0 +1,209 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 131072, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen2_5_VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null, + "use_fast": true +} diff --git a/posttrain_models/replan_qwen2_5_vl_7b/video_preprocessor_config.json b/posttrain_models/replan_qwen2_5_vl_7b/video_preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..b64d80b5644cc9753e2f7e348e1ac86a89e6e657 --- /dev/null +++ b/posttrain_models/replan_qwen2_5_vl_7b/video_preprocessor_config.json @@ -0,0 +1,86 @@ +{ + "_valid_kwargs_names": [ + "do_convert_rgb", + "do_resize", + "size", + "size_divisor", + "default_to_square", + "resample", + "do_rescale", + "rescale_factor", + "do_normalize", + "image_mean", + "image_std", + "do_pad", + "do_center_crop", + "crop_size", + "data_format", + "input_data_format", + "device", + "min_pixels", + "max_pixels", + "patch_size", + "temporal_patch_size", + "merge_size" + ], + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_pad": null, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.48145466, + 0.4578275, + 0.40821073 + ], + "image_processor_type": "Qwen2VLImageProcessor", + "image_std": [ + 0.26862954, + 0.26130258, + 0.27577711 + ], + "input_data_format": null, + "max_pixels": 12845056, + "merge_size": 2, + "min_pixels": 3136, + "model_valid_processing_keys": [ + "do_convert_rgb", + "do_resize", + "size", + "size_divisor", + "default_to_square", + "resample", + "do_rescale", + "rescale_factor", + "do_normalize", + "image_mean", + "image_std", + "do_pad", + "do_center_crop", + "crop_size", + "data_format", + "input_data_format", + "device", + "min_pixels", + "max_pixels", + "patch_size", + "temporal_patch_size", + "merge_size" + ], + "patch_size": 14, + "processor_class": "Qwen2_5_VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "longest_edge": 12845056, + "shortest_edge": 3136 + }, + "size_divisor": null, + "temporal_patch_size": 2, + "video_processor_type": "Qwen2VLVideoProcessor" +} diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v0-20260310-063316/args.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v0-20260310-063316/args.json new file mode 100644 index 0000000000000000000000000000000000000000..3f23f19f0f4a34c69b7305c9d71a7a35fd84ed4e --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v0-20260310-063316/args.json @@ -0,0 +1,389 @@ +{ + "output_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v0-20260310-063316", + "overwrite_output_dir": false, + "do_train": false, + "do_eval": false, + "do_predict": false, + "eval_strategy": "no", + "prediction_loss_only": false, + "per_device_train_batch_size": 4, + "per_device_eval_batch_size": 4, + "per_gpu_train_batch_size": null, + "per_gpu_eval_batch_size": null, + "gradient_accumulation_steps": 4, + "eval_accumulation_steps": null, + "eval_delay": 0, + "torch_empty_cache_steps": null, + "learning_rate": 1e-05, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "max_grad_norm": 1.0, + "num_train_epochs": 1.0, + "max_steps": -1, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_ratio": 0.03, + "warmup_steps": 0, + "log_level": "passive", + "log_level_replica": "warning", + "log_on_each_node": true, + "logging_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v0-20260310-063316/runs", + "logging_strategy": "steps", + "logging_first_step": true, + "logging_steps": 5, + "logging_nan_inf_filter": true, + "save_strategy": "epoch", + "save_steps": 500, + "save_total_limit": 1, + "save_safetensors": true, + "save_on_each_node": false, + "save_only_model": false, + "restore_callback_states_from_checkpoint": false, + "no_cuda": false, + "use_cpu": false, + "use_mps_device": false, + "seed": 42, + "data_seed": 42, + "jit_mode_eval": false, + "bf16": true, + "fp16": false, + "fp16_opt_level": "O1", + "half_precision_backend": "auto", + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": null, + "local_rank": 0, + "ddp_backend": null, + "tpu_num_cores": null, + "tpu_metrics_debug": false, + "debug": null, + "dataloader_drop_last": false, + "eval_steps": null, + "dataloader_num_workers": 24, + "dataloader_prefetch_factor": 2, + "past_index": -1, + "run_name": "sft_46k_multi_32k_timelens_1epoch_w_cot", + "disable_tqdm": null, + "remove_unused_columns": true, + "label_names": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "fsdp": [], + "fsdp_min_num_params": 0, + "fsdp_config": null, + "fsdp_transformer_layer_cls_to_wrap": null, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "deepspeed": { + "fp16": { + "enabled": "auto", + "loss_scale": 0, + "loss_scale_window": 1000, + "initial_scale_power": 16, + "hysteresis": 2, + "min_loss_scale": 1 + }, + "bf16": { + "enabled": "auto" + }, + "zero_optimization": { + "stage": 2, + "offload_optimizer": { + "device": "none", + "pin_memory": true + }, + "allgather_partitions": true, + "allgather_bucket_size": 200000000.0, + "overlap_comm": false, + "reduce_scatter": true, + "reduce_bucket_size": 200000000.0, + "contiguous_gradients": true + }, + "gradient_accumulation_steps": "auto", + "gradient_clipping": "auto", + "steps_per_print": 2000, + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "wall_clock_breakdown": false + }, + "label_smoothing_factor": 0.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "adafactor": false, + "group_by_length": false, + "length_column_name": "length", + "report_to": [ + "wandb" + ], + "project": "huggingface", + "trackio_space_id": "trackio", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": false, + "skip_memory_metrics": true, + "use_legacy_prediction_loop": false, + "push_to_hub": false, + "resume_from_checkpoint": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_token": null, + "hub_private_repo": null, + "hub_always_push": false, + "hub_revision": null, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "include_inputs_for_metrics": false, + "include_for_metrics": [], + "eval_do_concat_batches": true, + "fp16_backend": "auto", + "push_to_hub_model_id": null, + "push_to_hub_organization": null, + "push_to_hub_token": null, + "mp_parameters": "", + "auto_find_batch_size": false, + "full_determinism": false, + "torchdynamo": null, + "ray_scope": "last", + "ddp_timeout": 18000000, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "include_tokens_per_second": false, + "include_num_input_tokens_seen": false, + "neftune_noise_alpha": null, + "optim_target_modules": null, + "batch_eval_metrics": false, + "eval_on_start": false, + "use_liger_kernel": false, + "liger_kernel_config": null, + "eval_use_gather_object": false, + "average_tokens_across_devices": true, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": null, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "metric": [ + "temporal_grounding" + ], + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "Qwen/Qwen3-VL-4B-Instruct", + "model_type": "qwen3_vl", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": "flash_attn", + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_vl", + "system": null, + "max_length": 262144, + "truncation_strategy": "delete", + "max_pixels": null, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_free": false, + "padding_side": "right", + "loss_scale": "default", + "sequence_parallel_size": 4, + "response_prefix": null, + "template_backend": "swift", + "dataset": [ + "../data/cosmo-cap/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/moment10m/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/vtimellm/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/timelens-100k/train/train_w_cot_reasoning_32k.jsonl" + ], + "val_dataset": [], + "cached_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 1, + "load_from_cache_file": true, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "ckpt_dir": null, + "lora_modules": [], + "train_type": "full", + "adapters": [], + "external_plugins": [], + "model_kwargs": {}, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "custom_register_path": [], + "use_hf": true, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [ + "model.visual.merger", + "model.visual.deepstack_merger_list" + ], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": false, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": null, + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_lark_webhook_url": null, + "swanlab_lark_secret": null, + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "early_stop_interval": null, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3-VL-4B-Instruct", + "model_info": "ModelInfo(model_type='qwen3_vl', model_dir='/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling={'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default'}, is_moe_model=False, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_vl', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct', hf_model_id='Qwen/Qwen3-VL-2B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking', hf_model_id='Qwen/Qwen3-VL-2B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct', hf_model_id='Qwen/Qwen3-VL-4B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking', hf_model_id='Qwen/Qwen3-VL-4B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct', hf_model_id='Qwen/Qwen3-VL-8B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking', hf_model_id='Qwen/Qwen3-VL-8B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct', hf_model_id='Qwen/Qwen3-VL-32B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking', hf_model_id='Qwen/Qwen3-VL-32B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='qwen3_vl', get_function=, model_arch=MultiModelKeys(arch_name='qwen3_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model'], aligner=['model.visual.merger', 'model.visual.deepstack_merger_list'], vision_tower=['model.visual'], generator=[]), architectures=['Qwen3VLForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, is_reranker=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])", + "model_dir": "/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17", + "_val_dataset_exists": [], + "hub": "", + "evaluation_strategy": "epoch", + "training_args": "Seq2SeqTrainingArguments(output_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v0-20260310-063316', overwrite_output_dir=False, do_train=False, do_eval=False, do_predict=False, eval_strategy=, prediction_loss_only=False, per_device_train_batch_size=4, per_device_eval_batch_size=4, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=4, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=1e-05, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=1.0, max_steps=-1, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_ratio=0.03, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v0-20260310-063316/runs', logging_strategy=, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=, save_steps=500, save_total_limit=1, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, jit_mode_eval=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=None, dataloader_num_workers=24, dataloader_prefetch_factor=2, past_index=-1, run_name='sft_46k_multi_32k_timelens_1epoch_w_cot', disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, deepspeed={'fp16': {'enabled': 'auto', 'loss_scale': 0, 'loss_scale_window': 1000, 'initial_scale_power': 16, 'hysteresis': 2, 'min_loss_scale': 1}, 'bf16': {'enabled': 'auto'}, 'zero_optimization': {'stage': 2, 'offload_optimizer': {'device': 'none', 'pin_memory': True}, 'allgather_partitions': True, 'allgather_bucket_size': 200000000.0, 'overlap_comm': False, 'reduce_scatter': True, 'reduce_bucket_size': 200000000.0, 'contiguous_gradients': True}, 'gradient_accumulation_steps': 'auto', 'gradient_clipping': 'auto', 'steps_per_print': 2000, 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'wall_clock_breakdown': False}, label_smoothing_factor=0.0, optim=, optim_args=None, adafactor=False, group_by_length=False, length_column_name='length', report_to=['wandb'], project='huggingface', trackio_space_id='trackio', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint=None, hub_model_id=None, hub_strategy=, hub_token=None, hub_private_repo=None, hub_always_push=False, hub_revision=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=18000000, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, include_tokens_per_second=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, liger_kernel_config=None, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=True, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, check_model=True, acc_strategy='token', train_dataloader_shuffle=True, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, metric=['temporal_grounding'], eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, use_flash_ckpt=False, sft_alpha=0, chord_sft_dataset=[], chord_sft_per_device_train_batch_size=None, chord_enable_phi_function=False, chord_mu_warmup_steps=None, chord_mu_decay_steps=None, chord_mu_peak=None, chord_mu_valley=None, train_type='full', local_repo_path=None, galore_config=None, padding_side='right', padding_free=False, task_type='causal_lm', problem_type=None)" +} \ No newline at end of file diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v0-20260310-063316/logging.jsonl b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v0-20260310-063316/logging.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..6de7f9913328bb68ef91ff1dd171d15ba192bfcb --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v0-20260310-063316/logging.jsonl @@ -0,0 +1 @@ +{"model_parameter_info": "Qwen3VLForConditionalGeneration: 4437.8158M Params (4131.5732M Trainable [93.0993%]), 0.0001M Buffers.", "last_model_checkpoint": null, "best_model_checkpoint": null, "best_metric": null, "global_step": 0, "log_history": [], "memory": null} diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/args.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/args.json new file mode 100644 index 0000000000000000000000000000000000000000..e432bd237c39a8c90dd30db687c1feee7022e7bc --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/args.json @@ -0,0 +1,389 @@ +{ + "output_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639", + "overwrite_output_dir": false, + "do_train": false, + "do_eval": false, + "do_predict": false, + "eval_strategy": "no", + "prediction_loss_only": false, + "per_device_train_batch_size": 4, + "per_device_eval_batch_size": 4, + "per_gpu_train_batch_size": null, + "per_gpu_eval_batch_size": null, + "gradient_accumulation_steps": 4, + "eval_accumulation_steps": null, + "eval_delay": 0, + "torch_empty_cache_steps": null, + "learning_rate": 1e-05, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "max_grad_norm": 1.0, + "num_train_epochs": 1.0, + "max_steps": -1, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_ratio": 0.03, + "warmup_steps": 0, + "log_level": "passive", + "log_level_replica": "warning", + "log_on_each_node": true, + "logging_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/runs", + "logging_strategy": "steps", + "logging_first_step": true, + "logging_steps": 5, + "logging_nan_inf_filter": true, + "save_strategy": "epoch", + "save_steps": 500, + "save_total_limit": 1, + "save_safetensors": true, + "save_on_each_node": false, + "save_only_model": false, + "restore_callback_states_from_checkpoint": false, + "no_cuda": false, + "use_cpu": false, + "use_mps_device": false, + "seed": 42, + "data_seed": 42, + "jit_mode_eval": false, + "bf16": true, + "fp16": false, + "fp16_opt_level": "O1", + "half_precision_backend": "auto", + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": null, + "local_rank": 0, + "ddp_backend": null, + "tpu_num_cores": null, + "tpu_metrics_debug": false, + "debug": null, + "dataloader_drop_last": false, + "eval_steps": null, + "dataloader_num_workers": 24, + "dataloader_prefetch_factor": 2, + "past_index": -1, + "run_name": "sft_46k_multi_32k_timelens_1epoch_w_cot", + "disable_tqdm": null, + "remove_unused_columns": true, + "label_names": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "fsdp": [], + "fsdp_min_num_params": 0, + "fsdp_config": null, + "fsdp_transformer_layer_cls_to_wrap": null, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "deepspeed": { + "fp16": { + "enabled": "auto", + "loss_scale": 0, + "loss_scale_window": 1000, + "initial_scale_power": 16, + "hysteresis": 2, + "min_loss_scale": 1 + }, + "bf16": { + "enabled": "auto" + }, + "zero_optimization": { + "stage": 2, + "offload_optimizer": { + "device": "none", + "pin_memory": true + }, + "allgather_partitions": true, + "allgather_bucket_size": 200000000.0, + "overlap_comm": false, + "reduce_scatter": true, + "reduce_bucket_size": 200000000.0, + "contiguous_gradients": true + }, + "gradient_accumulation_steps": "auto", + "gradient_clipping": "auto", + "steps_per_print": 2000, + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "wall_clock_breakdown": false + }, + "label_smoothing_factor": 0.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "adafactor": false, + "group_by_length": false, + "length_column_name": "length", + "report_to": [ + "wandb" + ], + "project": "huggingface", + "trackio_space_id": "trackio", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": false, + "skip_memory_metrics": true, + "use_legacy_prediction_loop": false, + "push_to_hub": false, + "resume_from_checkpoint": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_token": null, + "hub_private_repo": null, + "hub_always_push": false, + "hub_revision": null, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "include_inputs_for_metrics": false, + "include_for_metrics": [], + "eval_do_concat_batches": true, + "fp16_backend": "auto", + "push_to_hub_model_id": null, + "push_to_hub_organization": null, + "push_to_hub_token": null, + "mp_parameters": "", + "auto_find_batch_size": false, + "full_determinism": false, + "torchdynamo": null, + "ray_scope": "last", + "ddp_timeout": 18000000, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "include_tokens_per_second": false, + "include_num_input_tokens_seen": false, + "neftune_noise_alpha": null, + "optim_target_modules": null, + "batch_eval_metrics": false, + "eval_on_start": false, + "use_liger_kernel": false, + "liger_kernel_config": null, + "eval_use_gather_object": false, + "average_tokens_across_devices": true, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": null, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "metric": [ + "temporal_grounding" + ], + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "Qwen/Qwen3-VL-4B-Instruct", + "model_type": "qwen3_vl", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": "flash_attn", + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_vl", + "system": null, + "max_length": 262144, + "truncation_strategy": "delete", + "max_pixels": null, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_free": false, + "padding_side": "right", + "loss_scale": "default", + "sequence_parallel_size": 4, + "response_prefix": null, + "template_backend": "swift", + "dataset": [ + "../data/cosmo-cap/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/moment10m/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/vtimellm/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/timelens-100k/train/train_w_cot_reasoning_32k.jsonl" + ], + "val_dataset": [], + "cached_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 1, + "load_from_cache_file": true, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "ckpt_dir": null, + "lora_modules": [], + "train_type": "full", + "adapters": [], + "external_plugins": [], + "model_kwargs": {}, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "custom_register_path": [], + "use_hf": true, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [ + "model.visual.merger", + "model.visual.deepstack_merger_list" + ], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": false, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": null, + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_lark_webhook_url": null, + "swanlab_lark_secret": null, + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "early_stop_interval": null, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3-VL-4B-Instruct", + "model_info": "ModelInfo(model_type='qwen3_vl', model_dir='/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling={'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default'}, is_moe_model=False, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_vl', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct', hf_model_id='Qwen/Qwen3-VL-2B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking', hf_model_id='Qwen/Qwen3-VL-2B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct', hf_model_id='Qwen/Qwen3-VL-4B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking', hf_model_id='Qwen/Qwen3-VL-4B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct', hf_model_id='Qwen/Qwen3-VL-8B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking', hf_model_id='Qwen/Qwen3-VL-8B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct', hf_model_id='Qwen/Qwen3-VL-32B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking', hf_model_id='Qwen/Qwen3-VL-32B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='qwen3_vl', get_function=, model_arch=MultiModelKeys(arch_name='qwen3_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model'], aligner=['model.visual.merger', 'model.visual.deepstack_merger_list'], vision_tower=['model.visual'], generator=[]), architectures=['Qwen3VLForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, is_reranker=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])", + "model_dir": "/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17", + "_val_dataset_exists": [], + "hub": "", + "evaluation_strategy": "epoch", + "training_args": "Seq2SeqTrainingArguments(output_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639', overwrite_output_dir=False, do_train=False, do_eval=False, do_predict=False, eval_strategy=, prediction_loss_only=False, per_device_train_batch_size=4, per_device_eval_batch_size=4, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=4, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=1e-05, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=1.0, max_steps=-1, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_ratio=0.03, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/runs', logging_strategy=, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=, save_steps=500, save_total_limit=1, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, jit_mode_eval=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=None, dataloader_num_workers=24, dataloader_prefetch_factor=2, past_index=-1, run_name='sft_46k_multi_32k_timelens_1epoch_w_cot', disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, deepspeed={'fp16': {'enabled': 'auto', 'loss_scale': 0, 'loss_scale_window': 1000, 'initial_scale_power': 16, 'hysteresis': 2, 'min_loss_scale': 1}, 'bf16': {'enabled': 'auto'}, 'zero_optimization': {'stage': 2, 'offload_optimizer': {'device': 'none', 'pin_memory': True}, 'allgather_partitions': True, 'allgather_bucket_size': 200000000.0, 'overlap_comm': False, 'reduce_scatter': True, 'reduce_bucket_size': 200000000.0, 'contiguous_gradients': True}, 'gradient_accumulation_steps': 'auto', 'gradient_clipping': 'auto', 'steps_per_print': 2000, 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'wall_clock_breakdown': False}, label_smoothing_factor=0.0, optim=, optim_args=None, adafactor=False, group_by_length=False, length_column_name='length', report_to=['wandb'], project='huggingface', trackio_space_id='trackio', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint=None, hub_model_id=None, hub_strategy=, hub_token=None, hub_private_repo=None, hub_always_push=False, hub_revision=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=18000000, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, include_tokens_per_second=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, liger_kernel_config=None, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=True, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, check_model=True, acc_strategy='token', train_dataloader_shuffle=True, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, metric=['temporal_grounding'], eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, use_flash_ckpt=False, sft_alpha=0, chord_sft_dataset=[], chord_sft_per_device_train_batch_size=None, chord_enable_phi_function=False, chord_mu_warmup_steps=None, chord_mu_decay_steps=None, chord_mu_peak=None, chord_mu_valley=None, train_type='full', local_repo_path=None, galore_config=None, padding_side='right', padding_free=False, task_type='causal_lm', problem_type=None)" +} \ No newline at end of file diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/added_tokens.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/added_tokens.json @@ -0,0 +1,28 @@ +{ + "
": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/args.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/args.json new file mode 100644 index 0000000000000000000000000000000000000000..e432bd237c39a8c90dd30db687c1feee7022e7bc --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/args.json @@ -0,0 +1,389 @@ +{ + "output_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639", + "overwrite_output_dir": false, + "do_train": false, + "do_eval": false, + "do_predict": false, + "eval_strategy": "no", + "prediction_loss_only": false, + "per_device_train_batch_size": 4, + "per_device_eval_batch_size": 4, + "per_gpu_train_batch_size": null, + "per_gpu_eval_batch_size": null, + "gradient_accumulation_steps": 4, + "eval_accumulation_steps": null, + "eval_delay": 0, + "torch_empty_cache_steps": null, + "learning_rate": 1e-05, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "max_grad_norm": 1.0, + "num_train_epochs": 1.0, + "max_steps": -1, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_ratio": 0.03, + "warmup_steps": 0, + "log_level": "passive", + "log_level_replica": "warning", + "log_on_each_node": true, + "logging_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/runs", + "logging_strategy": "steps", + "logging_first_step": true, + "logging_steps": 5, + "logging_nan_inf_filter": true, + "save_strategy": "epoch", + "save_steps": 500, + "save_total_limit": 1, + "save_safetensors": true, + "save_on_each_node": false, + "save_only_model": false, + "restore_callback_states_from_checkpoint": false, + "no_cuda": false, + "use_cpu": false, + "use_mps_device": false, + "seed": 42, + "data_seed": 42, + "jit_mode_eval": false, + "bf16": true, + "fp16": false, + "fp16_opt_level": "O1", + "half_precision_backend": "auto", + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": null, + "local_rank": 0, + "ddp_backend": null, + "tpu_num_cores": null, + "tpu_metrics_debug": false, + "debug": null, + "dataloader_drop_last": false, + "eval_steps": null, + "dataloader_num_workers": 24, + "dataloader_prefetch_factor": 2, + "past_index": -1, + "run_name": "sft_46k_multi_32k_timelens_1epoch_w_cot", + "disable_tqdm": null, + "remove_unused_columns": true, + "label_names": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "fsdp": [], + "fsdp_min_num_params": 0, + "fsdp_config": null, + "fsdp_transformer_layer_cls_to_wrap": null, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "deepspeed": { + "fp16": { + "enabled": "auto", + "loss_scale": 0, + "loss_scale_window": 1000, + "initial_scale_power": 16, + "hysteresis": 2, + "min_loss_scale": 1 + }, + "bf16": { + "enabled": "auto" + }, + "zero_optimization": { + "stage": 2, + "offload_optimizer": { + "device": "none", + "pin_memory": true + }, + "allgather_partitions": true, + "allgather_bucket_size": 200000000.0, + "overlap_comm": false, + "reduce_scatter": true, + "reduce_bucket_size": 200000000.0, + "contiguous_gradients": true + }, + "gradient_accumulation_steps": "auto", + "gradient_clipping": "auto", + "steps_per_print": 2000, + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "wall_clock_breakdown": false + }, + "label_smoothing_factor": 0.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "adafactor": false, + "group_by_length": false, + "length_column_name": "length", + "report_to": [ + "wandb" + ], + "project": "huggingface", + "trackio_space_id": "trackio", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": false, + "skip_memory_metrics": true, + "use_legacy_prediction_loop": false, + "push_to_hub": false, + "resume_from_checkpoint": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_token": null, + "hub_private_repo": null, + "hub_always_push": false, + "hub_revision": null, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "include_inputs_for_metrics": false, + "include_for_metrics": [], + "eval_do_concat_batches": true, + "fp16_backend": "auto", + "push_to_hub_model_id": null, + "push_to_hub_organization": null, + "push_to_hub_token": null, + "mp_parameters": "", + "auto_find_batch_size": false, + "full_determinism": false, + "torchdynamo": null, + "ray_scope": "last", + "ddp_timeout": 18000000, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "include_tokens_per_second": false, + "include_num_input_tokens_seen": false, + "neftune_noise_alpha": null, + "optim_target_modules": null, + "batch_eval_metrics": false, + "eval_on_start": false, + "use_liger_kernel": false, + "liger_kernel_config": null, + "eval_use_gather_object": false, + "average_tokens_across_devices": true, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": null, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "metric": [ + "temporal_grounding" + ], + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "Qwen/Qwen3-VL-4B-Instruct", + "model_type": "qwen3_vl", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": "flash_attn", + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_vl", + "system": null, + "max_length": 262144, + "truncation_strategy": "delete", + "max_pixels": null, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_free": false, + "padding_side": "right", + "loss_scale": "default", + "sequence_parallel_size": 4, + "response_prefix": null, + "template_backend": "swift", + "dataset": [ + "../data/cosmo-cap/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/moment10m/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/vtimellm/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/timelens-100k/train/train_w_cot_reasoning_32k.jsonl" + ], + "val_dataset": [], + "cached_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 1, + "load_from_cache_file": true, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "ckpt_dir": null, + "lora_modules": [], + "train_type": "full", + "adapters": [], + "external_plugins": [], + "model_kwargs": {}, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "custom_register_path": [], + "use_hf": true, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [ + "model.visual.merger", + "model.visual.deepstack_merger_list" + ], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": false, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": null, + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_lark_webhook_url": null, + "swanlab_lark_secret": null, + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "early_stop_interval": null, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3-VL-4B-Instruct", + "model_info": "ModelInfo(model_type='qwen3_vl', model_dir='/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling={'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default'}, is_moe_model=False, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_vl', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct', hf_model_id='Qwen/Qwen3-VL-2B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking', hf_model_id='Qwen/Qwen3-VL-2B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct', hf_model_id='Qwen/Qwen3-VL-4B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking', hf_model_id='Qwen/Qwen3-VL-4B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct', hf_model_id='Qwen/Qwen3-VL-8B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking', hf_model_id='Qwen/Qwen3-VL-8B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct', hf_model_id='Qwen/Qwen3-VL-32B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking', hf_model_id='Qwen/Qwen3-VL-32B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='qwen3_vl', get_function=, model_arch=MultiModelKeys(arch_name='qwen3_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model'], aligner=['model.visual.merger', 'model.visual.deepstack_merger_list'], vision_tower=['model.visual'], generator=[]), architectures=['Qwen3VLForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, is_reranker=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])", + "model_dir": "/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17", + "_val_dataset_exists": [], + "hub": "", + "evaluation_strategy": "epoch", + "training_args": "Seq2SeqTrainingArguments(output_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639', overwrite_output_dir=False, do_train=False, do_eval=False, do_predict=False, eval_strategy=, prediction_loss_only=False, per_device_train_batch_size=4, per_device_eval_batch_size=4, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=4, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=1e-05, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=1.0, max_steps=-1, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_ratio=0.03, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/runs', logging_strategy=, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=, save_steps=500, save_total_limit=1, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, jit_mode_eval=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=None, dataloader_num_workers=24, dataloader_prefetch_factor=2, past_index=-1, run_name='sft_46k_multi_32k_timelens_1epoch_w_cot', disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, deepspeed={'fp16': {'enabled': 'auto', 'loss_scale': 0, 'loss_scale_window': 1000, 'initial_scale_power': 16, 'hysteresis': 2, 'min_loss_scale': 1}, 'bf16': {'enabled': 'auto'}, 'zero_optimization': {'stage': 2, 'offload_optimizer': {'device': 'none', 'pin_memory': True}, 'allgather_partitions': True, 'allgather_bucket_size': 200000000.0, 'overlap_comm': False, 'reduce_scatter': True, 'reduce_bucket_size': 200000000.0, 'contiguous_gradients': True}, 'gradient_accumulation_steps': 'auto', 'gradient_clipping': 'auto', 'steps_per_print': 2000, 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'wall_clock_breakdown': False}, label_smoothing_factor=0.0, optim=, optim_args=None, adafactor=False, group_by_length=False, length_column_name='length', report_to=['wandb'], project='huggingface', trackio_space_id='trackio', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint=None, hub_model_id=None, hub_strategy=, hub_token=None, hub_private_repo=None, hub_always_push=False, hub_revision=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=18000000, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, include_tokens_per_second=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, liger_kernel_config=None, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=True, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, check_model=True, acc_strategy='token', train_dataloader_shuffle=True, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, metric=['temporal_grounding'], eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, use_flash_ckpt=False, sft_alpha=0, chord_sft_dataset=[], chord_sft_per_device_train_batch_size=None, chord_enable_phi_function=False, chord_mu_warmup_steps=None, chord_mu_decay_steps=None, chord_mu_peak=None, chord_mu_valley=None, train_type='full', local_repo_path=None, galore_config=None, padding_side='right', padding_free=False, task_type='causal_lm', problem_type=None)" +} \ No newline at end of file diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/chat_template.jinja b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/config.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/config.json new file mode 100644 index 0000000000000000000000000000000000000000..6cc71b039348c388053f379414d8d353760254f1 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/config.json @@ -0,0 +1,70 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "bfloat16", + "eos_token_id": 151645, + "hidden_size": 2560, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "bfloat16", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9728, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "pad_token_id": 151643, + "rms_norm_eps": 1e-06, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_type": "default" + }, + "rope_theta": 5000000, + "tie_word_embeddings": true, + "use_cache": false, + "vocab_size": 151936 + }, + "tie_word_embeddings": true, + "transformers_version": "4.57.0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 5, + 11, + 17 + ], + "depth": 24, + "dtype": "bfloat16", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "pad_token_id": 151643, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +} diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/generation_config.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..473d046ae2c09fde761a299ddc91ab5ba3cf67be --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 151643, + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "4.57.0" +} diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/latest b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/latest new file mode 100644 index 0000000000000000000000000000000000000000..e685ec613b50b37c63f121af9406583de0d8421c --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/latest @@ -0,0 +1 @@ +global_step2468 \ No newline at end of file diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/model.safetensors.index.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..5185ad4158d213752a144bd6053b766e0262d5f2 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/model.safetensors.index.json @@ -0,0 +1,722 @@ +{ + "metadata": { + "total_parameters": 4437815808, + "total_size": 9653543936 + }, + "weight_map": { + "lm_head.weight": "model-00002-of-00002.safetensors", + "model.language_model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.17.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.18.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.20.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.28.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.29.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.30.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.33.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.34.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.input_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_norm.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.35.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "model.language_model.norm.weight": "model-00002-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00001-of-00002.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.0.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.1.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.deepstack_merger_list.2.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00001-of-00002.safetensors", + "model.visual.merger.norm.bias": "model-00001-of-00002.safetensors", + "model.visual.merger.norm.weight": "model-00001-of-00002.safetensors", + "model.visual.patch_embed.proj.bias": "model-00001-of-00002.safetensors", + "model.visual.patch_embed.proj.weight": "model-00001-of-00002.safetensors", + "model.visual.pos_embed.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/preprocessor_config.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..2ea84a437d448ff71b08df68fdd949d5cc4ebb64 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/preprocessor_config.json @@ -0,0 +1,21 @@ +{ + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "patch_size": 16, + "temporal_patch_size": 2, + "merge_size": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "processor_class": "Qwen3VLProcessor", + "image_processor_type": "Qwen2VLImageProcessorFast" +} \ No newline at end of file diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_0.pth b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..4b0bea5e7a82c1886ccb79d111257405625494a6 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f8f92cf63e0989759370d24108b469c492c12202403f036015307ce49f12cedc +size 16389 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_1.pth b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..2c94d69496797fa9c0f83f8a534c8ba692cb143a --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9ed40a0a4e9f365d2c6cc004d97e6705894eba46c8be4c160c1455bc3062dee1 +size 16389 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_2.pth b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..e3d2f614c50841e8255840051884b16a6e22c38a --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d688b304d19c260b5cfa471535ed51d7e1d60b3a0d0159dfd1a04b87904a9f42 +size 16389 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_3.pth b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_3.pth new file mode 100644 index 0000000000000000000000000000000000000000..84660ff39da893f3edf58f4a54971ca424b5ae70 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_3.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9967425ebcaee80d9b518fa0244d52f739b1b983d87cda71d5fede0c073e9d3b +size 16389 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_4.pth b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_4.pth new file mode 100644 index 0000000000000000000000000000000000000000..fd63d0f91d990383a77a110da21707c2e4161995 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_4.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:469900fd39c667ffbd49c3c407c0ba317a1e9f5f9339a99b5d38423b7d0ce6d4 +size 16389 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_5.pth b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_5.pth new file mode 100644 index 0000000000000000000000000000000000000000..a9f2a80a12ac3102dc70232f90338ece968e5537 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_5.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:124688471ff2a6e80f2fcefedbf741fb18d08dd539d5bd07a52e81be545142a5 +size 16389 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_6.pth b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_6.pth new file mode 100644 index 0000000000000000000000000000000000000000..ffdeaa99e14137e82d32d685f2311f4ac57f4cd1 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_6.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1e69f1ced9f992a72c948698e5eb06088610788988cdb2fdbdd624e064319d60 +size 16389 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_7.pth b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_7.pth new file mode 100644 index 0000000000000000000000000000000000000000..ad30a51583bfc3be86c6b6aa3bc2e815019b3e77 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/rng_state_7.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a376268a55d6ee10c371c06aa952334c4c6a1af9ea2d71b1951a57367a0c6722 +size 16389 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/scheduler.pt b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..0348713cdb41c46c96b3c98b8d38c45701abc525 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3a5e73997c8f3361122e8333e862f8e79c5127aeb1716c3c59bebbe08a7f6d98 +size 1465 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/special_tokens_map.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..ac23c0aaa2434523c494330aeb79c58395378103 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/special_tokens_map.json @@ -0,0 +1,31 @@ +{ + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "eos_token": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/tokenizer_config.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..fec7f182135c384a08ea2fc9635566ca79c45afc --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/tokenizer_config.json @@ -0,0 +1,240 @@ +{ + "add_bos_token": false, + "add_prefix_space": false, + "added_tokens_decoder": { + "151643": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151644": { + "content": "<|im_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151645": { + "content": "<|im_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151646": { + "content": "<|object_ref_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151647": { + "content": "<|object_ref_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151648": { + "content": "<|box_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151649": { + "content": "<|box_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151650": { + "content": "<|quad_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151651": { + "content": "<|quad_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151652": { + "content": "<|vision_start|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151653": { + "content": "<|vision_end|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151654": { + "content": "<|vision_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151655": { + "content": "<|image_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151656": { + "content": "<|video_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "151657": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151658": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151659": { + "content": "<|fim_prefix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151660": { + "content": "<|fim_middle|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151661": { + "content": "<|fim_suffix|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151662": { + "content": "<|fim_pad|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151663": { + "content": "<|repo_name|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151664": { + "content": "<|file_sep|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151665": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151666": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151667": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151668": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "<|im_start|>", + "<|im_end|>", + "<|object_ref_start|>", + "<|object_ref_end|>", + "<|box_start|>", + "<|box_end|>", + "<|quad_start|>", + "<|quad_end|>", + "<|vision_start|>", + "<|vision_end|>", + "<|vision_pad|>", + "<|image_pad|>", + "<|video_pad|>" + ], + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_special_tokens": {}, + "model_max_length": 262144, + "pad_token": "<|endoftext|>", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/trainer_state.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..2e8026993aa6591c4fd7fc6de9bbd8e640c75b49 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/trainer_state.json @@ -0,0 +1,5962 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 2468, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0004053095551727632, + "grad_norm": 1.2059801816940308, + "learning_rate": 1.3333333333333336e-07, + "loss": 0.703125, + "step": 1, + "tIoU": 0.2516516149044037, + "tRecall_30": 0.7859375476837158, + "tRecall_50": 0.5697917342185974, + "tRecall_70": 0.5463542342185974, + "token_acc": 0.8105131343028716 + }, + { + "epoch": 0.002026547775863816, + "grad_norm": 1.4883641004562378, + "learning_rate": 6.666666666666667e-07, + "loss": 0.733642578125, + "step": 5, + "tIoU": 0.2318701297044754, + "tRecall_30": 0.7443870902061462, + "tRecall_50": 0.6183698773384094, + "tRecall_70": 0.5276147723197937, + "token_acc": 0.8065031164604378 + }, + { + "epoch": 0.004053095551727632, + "grad_norm": 1.1400638818740845, + "learning_rate": 1.3333333333333334e-06, + "loss": 0.74443359375, + "step": 10, + "tIoU": 0.252414608001709, + "tRecall_30": 0.7867266654968261, + "tRecall_50": 0.6461015701293945, + "tRecall_70": 0.5109973907470703, + "token_acc": 0.8026641380519708 + }, + { + "epoch": 0.006079643327591448, + "grad_norm": 0.8875113129615784, + "learning_rate": 2.0000000000000003e-06, + "loss": 0.675390625, + "step": 15, + "tIoU": 0.2845264196395874, + "tRecall_30": 0.8465773582458496, + "tRecall_50": 0.7617856979370117, + "tRecall_70": 0.7227232456207275, + "token_acc": 0.8118741947509205 + }, + { + "epoch": 0.008106191103455264, + "grad_norm": 1.0106897354125977, + "learning_rate": 2.666666666666667e-06, + "loss": 0.573974609375, + "step": 20, + "tIoU": 0.24748537540435792, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8294778755378588 + }, + { + "epoch": 0.01013273887931908, + "grad_norm": 0.7312051057815552, + "learning_rate": 3.3333333333333333e-06, + "loss": 0.474853515625, + "step": 25, + "tIoU": 0.8458622932434082, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.852569836141382 + }, + { + "epoch": 0.012159286655182896, + "grad_norm": 0.48748230934143066, + "learning_rate": 4.000000000000001e-06, + "loss": 0.422412109375, + "step": 30, + "tIoU": 0.9924302101135254, + "tRecall_30": 0.9893749237060547, + "tRecall_50": 0.9893749237060547, + "tRecall_70": 0.9893749237060547, + "token_acc": 0.8628151138194027 + }, + { + "epoch": 0.014185834431046712, + "grad_norm": 0.5568686723709106, + "learning_rate": 4.666666666666667e-06, + "loss": 0.4025146484375, + "step": 35, + "tIoU": 0.9956202507019043, + "tRecall_30": 0.9938838958740235, + "tRecall_50": 0.9938838958740235, + "tRecall_70": 0.9938838958740235, + "token_acc": 0.8690114445460885 + }, + { + "epoch": 0.016212382206910528, + "grad_norm": 0.6236920356750488, + "learning_rate": 5.333333333333334e-06, + "loss": 0.4341064453125, + "step": 40, + "tIoU": 0.9995620727539063, + "tRecall_30": 0.9987499237060546, + "tRecall_50": 0.9987499237060546, + "tRecall_70": 0.9987499237060546, + "token_acc": 0.8577157247139078 + }, + { + "epoch": 0.018238929982774345, + "grad_norm": 0.6035606861114502, + "learning_rate": 6e-06, + "loss": 0.3864501953125, + "step": 45, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8727699530516432 + }, + { + "epoch": 0.02026547775863816, + "grad_norm": 0.4480346739292145, + "learning_rate": 6.666666666666667e-06, + "loss": 0.3807373046875, + "step": 50, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8747324914642315 + }, + { + "epoch": 0.022292025534501977, + "grad_norm": 0.5582020282745361, + "learning_rate": 7.333333333333333e-06, + "loss": 0.420947265625, + "step": 55, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8620249574588769 + }, + { + "epoch": 0.02431857331036579, + "grad_norm": 0.6299155950546265, + "learning_rate": 8.000000000000001e-06, + "loss": 0.3962890625, + "step": 60, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8698634443214969 + }, + { + "epoch": 0.02634512108622961, + "grad_norm": 0.3295726478099823, + "learning_rate": 8.666666666666668e-06, + "loss": 0.361181640625, + "step": 65, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8810466902481946 + }, + { + "epoch": 0.028371668862093424, + "grad_norm": 0.5453260540962219, + "learning_rate": 9.333333333333334e-06, + "loss": 0.42099609375, + "step": 70, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8612275187707351 + }, + { + "epoch": 0.03039821663795724, + "grad_norm": 0.3517334461212158, + "learning_rate": 1e-05, + "loss": 0.3656982421875, + "step": 75, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8824227347762289 + }, + { + "epoch": 0.032424764413821056, + "grad_norm": 0.6326484084129333, + "learning_rate": 9.999892280877701e-06, + "loss": 0.405224609375, + "step": 80, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.866988653320259 + }, + { + "epoch": 0.03445131218968487, + "grad_norm": 0.4170985221862793, + "learning_rate": 9.999569128152163e-06, + "loss": 0.374755859375, + "step": 85, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8817892842977771 + }, + { + "epoch": 0.03647785996554869, + "grad_norm": 0.6364180445671082, + "learning_rate": 9.99903055574728e-06, + "loss": 0.3975830078125, + "step": 90, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8686657147390132 + }, + { + "epoch": 0.038504407741412505, + "grad_norm": 0.42150235176086426, + "learning_rate": 9.99827658686887e-06, + "loss": 0.386767578125, + "step": 95, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.876250368819027 + }, + { + "epoch": 0.04053095551727632, + "grad_norm": 0.550065279006958, + "learning_rate": 9.997307254003679e-06, + "loss": 0.3949951171875, + "step": 100, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8704353888079093 + }, + { + "epoch": 0.042557503293140134, + "grad_norm": 0.419331431388855, + "learning_rate": 9.99612259891798e-06, + "loss": 0.3871826171875, + "step": 105, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8742890360525686 + }, + { + "epoch": 0.044584051069003955, + "grad_norm": 0.4708976447582245, + "learning_rate": 9.99472267265578e-06, + "loss": 0.396240234375, + "step": 110, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8678600939563964 + }, + { + "epoch": 0.04661059884486777, + "grad_norm": 0.4568882882595062, + "learning_rate": 9.993107535536606e-06, + "loss": 0.392578125, + "step": 115, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8693056473022323 + }, + { + "epoch": 0.04863714662073158, + "grad_norm": 0.46889740228652954, + "learning_rate": 9.99127725715292e-06, + "loss": 0.375634765625, + "step": 120, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8766478017806612 + }, + { + "epoch": 0.0506636943965954, + "grad_norm": 0.4288806617259979, + "learning_rate": 9.989231916367114e-06, + "loss": 0.379248046875, + "step": 125, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8759755899858728 + }, + { + "epoch": 0.05269024217245922, + "grad_norm": 0.460036039352417, + "learning_rate": 9.986971601308115e-06, + "loss": 0.4231689453125, + "step": 130, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8616542750929368 + }, + { + "epoch": 0.05471678994832303, + "grad_norm": 0.5301831364631653, + "learning_rate": 9.984496409367586e-06, + "loss": 0.36162109375, + "step": 135, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8786379884342778 + }, + { + "epoch": 0.05674333772418685, + "grad_norm": 0.47949114441871643, + "learning_rate": 9.981806447195723e-06, + "loss": 0.4042724609375, + "step": 140, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8658658136974522 + }, + { + "epoch": 0.05876988550005066, + "grad_norm": 0.4840526282787323, + "learning_rate": 9.978901830696678e-06, + "loss": 0.368505859375, + "step": 145, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8775575871141142 + }, + { + "epoch": 0.06079643327591448, + "grad_norm": 0.40476393699645996, + "learning_rate": 9.975782685023543e-06, + "loss": 0.3725830078125, + "step": 150, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.877293148093681 + }, + { + "epoch": 0.06282298105177829, + "grad_norm": 0.425466388463974, + "learning_rate": 9.972449144572973e-06, + "loss": 0.37080078125, + "step": 155, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8802019781715128 + }, + { + "epoch": 0.06484952882764211, + "grad_norm": 0.560032069683075, + "learning_rate": 9.968901352979388e-06, + "loss": 0.4077880859375, + "step": 160, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8662045556516481 + }, + { + "epoch": 0.06687607660350593, + "grad_norm": 0.37908798456192017, + "learning_rate": 9.965139463108788e-06, + "loss": 0.3838623046875, + "step": 165, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8728647211558769 + }, + { + "epoch": 0.06890262437936974, + "grad_norm": 0.49446025490760803, + "learning_rate": 9.961163637052161e-06, + "loss": 0.4210205078125, + "step": 170, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.860889583704378 + }, + { + "epoch": 0.07092917215523356, + "grad_norm": 0.48118704557418823, + "learning_rate": 9.956974046118506e-06, + "loss": 0.377734375, + "step": 175, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8757897836007655 + }, + { + "epoch": 0.07295571993109738, + "grad_norm": 0.5156424045562744, + "learning_rate": 9.952570870827444e-06, + "loss": 0.413671875, + "step": 180, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8644025372652602 + }, + { + "epoch": 0.07498226770696119, + "grad_norm": 0.37811586260795593, + "learning_rate": 9.94795430090145e-06, + "loss": 0.38076171875, + "step": 185, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8748112623583383 + }, + { + "epoch": 0.07700881548282501, + "grad_norm": 0.42654335498809814, + "learning_rate": 9.943124535257665e-06, + "loss": 0.3935302734375, + "step": 190, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8720547928555367 + }, + { + "epoch": 0.07903536325868882, + "grad_norm": 0.43498939275741577, + "learning_rate": 9.938081781999337e-06, + "loss": 0.3921142578125, + "step": 195, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8706406149503881 + }, + { + "epoch": 0.08106191103455264, + "grad_norm": 0.4303409159183502, + "learning_rate": 9.932826258406847e-06, + "loss": 0.37509765625, + "step": 200, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8761669059603685 + }, + { + "epoch": 0.08308845881041646, + "grad_norm": 0.5523289442062378, + "learning_rate": 9.927358190928349e-06, + "loss": 0.391943359375, + "step": 205, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8704423086005763 + }, + { + "epoch": 0.08511500658628027, + "grad_norm": 0.4943441152572632, + "learning_rate": 9.921677815170018e-06, + "loss": 0.399560546875, + "step": 210, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8697048980833587 + }, + { + "epoch": 0.08714155436214409, + "grad_norm": 0.506697416305542, + "learning_rate": 9.91578537588589e-06, + "loss": 0.4166015625, + "step": 215, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8614923144262211 + }, + { + "epoch": 0.08916810213800791, + "grad_norm": 0.43591195344924927, + "learning_rate": 9.909681126967317e-06, + "loss": 0.39150390625, + "step": 220, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8727103654049425 + }, + { + "epoch": 0.09119464991387172, + "grad_norm": 0.5096474289894104, + "learning_rate": 9.903365331432037e-06, + "loss": 0.383642578125, + "step": 225, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8732293534195278 + }, + { + "epoch": 0.09322119768973554, + "grad_norm": 0.4196990132331848, + "learning_rate": 9.896838261412828e-06, + "loss": 0.422021484375, + "step": 230, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8624010501078235 + }, + { + "epoch": 0.09524774546559935, + "grad_norm": 0.5218161940574646, + "learning_rate": 9.890100198145793e-06, + "loss": 0.3818115234375, + "step": 235, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8751534205991717 + }, + { + "epoch": 0.09727429324146317, + "grad_norm": 0.37742888927459717, + "learning_rate": 9.883151431958236e-06, + "loss": 0.392919921875, + "step": 240, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8700355812097611 + }, + { + "epoch": 0.09930084101732699, + "grad_norm": 0.46909821033477783, + "learning_rate": 9.875992262256156e-06, + "loss": 0.39609375, + "step": 245, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8688558903912973 + }, + { + "epoch": 0.1013273887931908, + "grad_norm": 0.6484021544456482, + "learning_rate": 9.868622997511341e-06, + "loss": 0.3753662109375, + "step": 250, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8763179237631792 + }, + { + "epoch": 0.10335393656905462, + "grad_norm": 0.5185469388961792, + "learning_rate": 9.861043955248085e-06, + "loss": 0.3893798828125, + "step": 255, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8709359037257222 + }, + { + "epoch": 0.10538048434491844, + "grad_norm": 0.48241207003593445, + "learning_rate": 9.853255462029501e-06, + "loss": 0.401123046875, + "step": 260, + "tIoU": 0.9969846725463867, + "tRecall_30": 0.9979166030883789, + "tRecall_50": 0.9979166030883789, + "tRecall_70": 0.9979166030883789, + "token_acc": 0.86794682422452 + }, + { + "epoch": 0.10740703212078224, + "grad_norm": 0.4943467676639557, + "learning_rate": 9.845257853443448e-06, + "loss": 0.4021484375, + "step": 265, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8677978955351218 + }, + { + "epoch": 0.10943357989664607, + "grad_norm": 0.5944103002548218, + "learning_rate": 9.83705147408808e-06, + "loss": 0.3967041015625, + "step": 270, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8712227138360563 + }, + { + "epoch": 0.11146012767250987, + "grad_norm": 0.4518413245677948, + "learning_rate": 9.828636677556987e-06, + "loss": 0.368603515625, + "step": 275, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8792639007674004 + }, + { + "epoch": 0.1134866754483737, + "grad_norm": 0.4986145496368408, + "learning_rate": 9.820013826423966e-06, + "loss": 0.408349609375, + "step": 280, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8656716417910447 + }, + { + "epoch": 0.11551322322423752, + "grad_norm": 0.418460488319397, + "learning_rate": 9.811183292227406e-06, + "loss": 0.39873046875, + "step": 285, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8686417924266272 + }, + { + "epoch": 0.11753977100010132, + "grad_norm": 0.5732330083847046, + "learning_rate": 9.802145455454258e-06, + "loss": 0.407666015625, + "step": 290, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8654167783778263 + }, + { + "epoch": 0.11956631877596514, + "grad_norm": 0.42870670557022095, + "learning_rate": 9.792900705523661e-06, + "loss": 0.392236328125, + "step": 295, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8706978203295145 + }, + { + "epoch": 0.12159286655182897, + "grad_norm": 0.7802377939224243, + "learning_rate": 9.783449440770156e-06, + "loss": 0.417626953125, + "step": 300, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8623496986015365 + }, + { + "epoch": 0.12361941432769277, + "grad_norm": 0.552323043346405, + "learning_rate": 9.773792068426523e-06, + "loss": 0.389306640625, + "step": 305, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8697585808039884 + }, + { + "epoch": 0.12564596210355658, + "grad_norm": 0.48368698358535767, + "learning_rate": 9.763929004606224e-06, + "loss": 0.402001953125, + "step": 310, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8682366737867447 + }, + { + "epoch": 0.12767250987942041, + "grad_norm": 0.3376498520374298, + "learning_rate": 9.753860674285496e-06, + "loss": 0.37734375, + "step": 315, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8767554365694424 + }, + { + "epoch": 0.12969905765528422, + "grad_norm": 0.26270627975463867, + "learning_rate": 9.743587511285017e-06, + "loss": 0.376708984375, + "step": 320, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8774723504125398 + }, + { + "epoch": 0.13172560543114803, + "grad_norm": 0.39964982867240906, + "learning_rate": 9.733109958251232e-06, + "loss": 0.381787109375, + "step": 325, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8723410804986917 + }, + { + "epoch": 0.13375215320701186, + "grad_norm": 0.4558398425579071, + "learning_rate": 9.722428466637266e-06, + "loss": 0.359912109375, + "step": 330, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8835972975966331 + }, + { + "epoch": 0.13577870098287567, + "grad_norm": 0.6383611559867859, + "learning_rate": 9.711543496683477e-06, + "loss": 0.417822265625, + "step": 335, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8618028783968806 + }, + { + "epoch": 0.13780524875873948, + "grad_norm": 0.6163831353187561, + "learning_rate": 9.700455517397633e-06, + "loss": 0.413818359375, + "step": 340, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8630407959419495 + }, + { + "epoch": 0.13983179653460331, + "grad_norm": 0.5585286021232605, + "learning_rate": 9.68916500653469e-06, + "loss": 0.4010986328125, + "step": 345, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8668367878346006 + }, + { + "epoch": 0.14185834431046712, + "grad_norm": 0.532497763633728, + "learning_rate": 9.677672450576214e-06, + "loss": 0.4055908203125, + "step": 350, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8656914720402236 + }, + { + "epoch": 0.14388489208633093, + "grad_norm": 0.5201588869094849, + "learning_rate": 9.665978344709428e-06, + "loss": 0.373193359375, + "step": 355, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8762551519158639 + }, + { + "epoch": 0.14591143986219476, + "grad_norm": 0.39076343178749084, + "learning_rate": 9.654083192805854e-06, + "loss": 0.39013671875, + "step": 360, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8710004305063417 + }, + { + "epoch": 0.14793798763805857, + "grad_norm": 0.4675477147102356, + "learning_rate": 9.641987507399623e-06, + "loss": 0.3804931640625, + "step": 365, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8750331619439847 + }, + { + "epoch": 0.14996453541392238, + "grad_norm": 0.5155381560325623, + "learning_rate": 9.629691809665385e-06, + "loss": 0.3907958984375, + "step": 370, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8691968844244972 + }, + { + "epoch": 0.15199108318978619, + "grad_norm": 0.4530424475669861, + "learning_rate": 9.61719662939584e-06, + "loss": 0.402197265625, + "step": 375, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8673240255194368 + }, + { + "epoch": 0.15401763096565002, + "grad_norm": 0.4631502330303192, + "learning_rate": 9.604502504978934e-06, + "loss": 0.39306640625, + "step": 380, + "tIoU": 0.9999645233154297, + "tRecall_30": 1.0, + "tRecall_50": 0.9998046875, + "tRecall_70": 0.9998046875, + "token_acc": 0.8692022617865057 + }, + { + "epoch": 0.15604417874151383, + "grad_norm": 0.45243507623672485, + "learning_rate": 9.591609983374641e-06, + "loss": 0.3896240234375, + "step": 385, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.871999049987795 + }, + { + "epoch": 0.15807072651737764, + "grad_norm": 0.3028954863548279, + "learning_rate": 9.578519620091406e-06, + "loss": 0.3758544921875, + "step": 390, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8807784330048143 + }, + { + "epoch": 0.16009727429324147, + "grad_norm": 0.42049020528793335, + "learning_rate": 9.565231979162209e-06, + "loss": 0.36181640625, + "step": 395, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8822334418785289 + }, + { + "epoch": 0.16212382206910528, + "grad_norm": 0.5457665324211121, + "learning_rate": 9.551747633120252e-06, + "loss": 0.3998291015625, + "step": 400, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8688331201649054 + }, + { + "epoch": 0.16415036984496909, + "grad_norm": 0.5921686887741089, + "learning_rate": 9.538067162974307e-06, + "loss": 0.404541015625, + "step": 405, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8673737888832228 + }, + { + "epoch": 0.16617691762083292, + "grad_norm": 0.44998884201049805, + "learning_rate": 9.524191158183667e-06, + "loss": 0.4027587890625, + "step": 410, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8659621695270872 + }, + { + "epoch": 0.16820346539669673, + "grad_norm": 0.4771840274333954, + "learning_rate": 9.510120216632756e-06, + "loss": 0.395751953125, + "step": 415, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8684010315925209 + }, + { + "epoch": 0.17023001317256053, + "grad_norm": 0.4623829126358032, + "learning_rate": 9.495854944605365e-06, + "loss": 0.3904541015625, + "step": 420, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8688424574821375 + }, + { + "epoch": 0.17225656094842437, + "grad_norm": 0.38424453139305115, + "learning_rate": 9.481395956758524e-06, + "loss": 0.4014892578125, + "step": 425, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8684298649367362 + }, + { + "epoch": 0.17428310872428818, + "grad_norm": 0.46710488200187683, + "learning_rate": 9.466743876096025e-06, + "loss": 0.3872802734375, + "step": 430, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8724153877983181 + }, + { + "epoch": 0.17630965650015198, + "grad_norm": 0.5199359059333801, + "learning_rate": 9.451899333941576e-06, + "loss": 0.3877197265625, + "step": 435, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8741814181929807 + }, + { + "epoch": 0.17833620427601582, + "grad_norm": 0.4269919693470001, + "learning_rate": 9.4368629699116e-06, + "loss": 0.387841796875, + "step": 440, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8710228758766247 + }, + { + "epoch": 0.18036275205187963, + "grad_norm": 0.4944726526737213, + "learning_rate": 9.421635431887669e-06, + "loss": 0.37880859375, + "step": 445, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8753988453659445 + }, + { + "epoch": 0.18238929982774343, + "grad_norm": 0.5226631164550781, + "learning_rate": 9.406217375988597e-06, + "loss": 0.3662109375, + "step": 450, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.881288582122199 + }, + { + "epoch": 0.18441584760360724, + "grad_norm": 0.4460074007511139, + "learning_rate": 9.390609466542162e-06, + "loss": 0.3919921875, + "step": 455, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8687119839296501 + }, + { + "epoch": 0.18644239537947108, + "grad_norm": 0.41100502014160156, + "learning_rate": 9.374812376056486e-06, + "loss": 0.395458984375, + "step": 460, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8694572158365261 + }, + { + "epoch": 0.18846894315533488, + "grad_norm": 0.43790924549102783, + "learning_rate": 9.35882678519106e-06, + "loss": 0.3611083984375, + "step": 465, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8846521273125495 + }, + { + "epoch": 0.1904954909311987, + "grad_norm": 0.447274386882782, + "learning_rate": 9.34265338272741e-06, + "loss": 0.3850341796875, + "step": 470, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8730763461826908 + }, + { + "epoch": 0.19252203870706253, + "grad_norm": 0.4919845163822174, + "learning_rate": 9.326292865539423e-06, + "loss": 0.3983154296875, + "step": 475, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8671222277848937 + }, + { + "epoch": 0.19454858648292633, + "grad_norm": 0.4264659881591797, + "learning_rate": 9.309745938563322e-06, + "loss": 0.37763671875, + "step": 480, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8753093047507976 + }, + { + "epoch": 0.19657513425879014, + "grad_norm": 0.43917012214660645, + "learning_rate": 9.293013314767284e-06, + "loss": 0.4005615234375, + "step": 485, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8670107159717776 + }, + { + "epoch": 0.19860168203465398, + "grad_norm": 0.398868203163147, + "learning_rate": 9.276095715120727e-06, + "loss": 0.3945556640625, + "step": 490, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8710031307739684 + }, + { + "epoch": 0.20062822981051778, + "grad_norm": 0.45519566535949707, + "learning_rate": 9.25899386856325e-06, + "loss": 0.378173828125, + "step": 495, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8744337476399202 + }, + { + "epoch": 0.2026547775863816, + "grad_norm": 0.4370631277561188, + "learning_rate": 9.241708511973214e-06, + "loss": 0.3460693359375, + "step": 500, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8840001404214985 + }, + { + "epoch": 0.20468132536224543, + "grad_norm": 0.3996580243110657, + "learning_rate": 9.224240390135988e-06, + "loss": 0.364599609375, + "step": 505, + "tIoU": 0.9998649597167969, + "tRecall_30": 0.9998568534851074, + "tRecall_50": 0.9998568534851074, + "tRecall_70": 0.9998568534851074, + "token_acc": 0.8795211009067047 + }, + { + "epoch": 0.20670787313810923, + "grad_norm": 0.48396146297454834, + "learning_rate": 9.20659025571188e-06, + "loss": 0.3871826171875, + "step": 510, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8723862355459301 + }, + { + "epoch": 0.20873442091397304, + "grad_norm": 0.5639007091522217, + "learning_rate": 9.188758869203682e-06, + "loss": 0.3907958984375, + "step": 515, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8702464822224618 + }, + { + "epoch": 0.21076096868983688, + "grad_norm": 0.46063733100891113, + "learning_rate": 9.170746998923918e-06, + "loss": 0.3768310546875, + "step": 520, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8767145893566742 + }, + { + "epoch": 0.21278751646570068, + "grad_norm": 0.4143420457839966, + "learning_rate": 9.152555420961729e-06, + "loss": 0.3778564453125, + "step": 525, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8748213349849921 + }, + { + "epoch": 0.2148140642415645, + "grad_norm": 0.5685514211654663, + "learning_rate": 9.13418491914944e-06, + "loss": 0.406494140625, + "step": 530, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8642833706814105 + }, + { + "epoch": 0.21684061201742832, + "grad_norm": 0.5420283079147339, + "learning_rate": 9.115636285028784e-06, + "loss": 0.3872314453125, + "step": 535, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8724378376307421 + }, + { + "epoch": 0.21886715979329213, + "grad_norm": 0.6400812864303589, + "learning_rate": 9.096910317816796e-06, + "loss": 0.4113037109375, + "step": 540, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8705328273096042 + }, + { + "epoch": 0.22089370756915594, + "grad_norm": 0.689870297908783, + "learning_rate": 9.078007824371376e-06, + "loss": 0.40126953125, + "step": 545, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8671837458878032 + }, + { + "epoch": 0.22292025534501975, + "grad_norm": 0.4439537227153778, + "learning_rate": 9.058929619156527e-06, + "loss": 0.3592529296875, + "step": 550, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8815546864532758 + }, + { + "epoch": 0.22494680312088358, + "grad_norm": 0.38134971261024475, + "learning_rate": 9.039676524207255e-06, + "loss": 0.37060546875, + "step": 555, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8778313570172185 + }, + { + "epoch": 0.2269733508967474, + "grad_norm": 0.3596774637699127, + "learning_rate": 9.020249369094159e-06, + "loss": 0.40673828125, + "step": 560, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8652295664267323 + }, + { + "epoch": 0.2289998986726112, + "grad_norm": 0.6109125018119812, + "learning_rate": 9.000648990887674e-06, + "loss": 0.3809326171875, + "step": 565, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8743921025504816 + }, + { + "epoch": 0.23102644644847503, + "grad_norm": 0.47461310029029846, + "learning_rate": 8.980876234122017e-06, + "loss": 0.3836181640625, + "step": 570, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.874513549428124 + }, + { + "epoch": 0.23305299422433884, + "grad_norm": 0.41040199995040894, + "learning_rate": 8.960931950758792e-06, + "loss": 0.347021484375, + "step": 575, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8863791737653685 + }, + { + "epoch": 0.23507954200020265, + "grad_norm": 0.4203317165374756, + "learning_rate": 8.940817000150273e-06, + "loss": 0.3716064453125, + "step": 580, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8802946802696302 + }, + { + "epoch": 0.23710608977606648, + "grad_norm": 0.446027934551239, + "learning_rate": 8.920532249002394e-06, + "loss": 0.390380859375, + "step": 585, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8720234062211272 + }, + { + "epoch": 0.2391326375519303, + "grad_norm": 0.9307211637496948, + "learning_rate": 8.900078571337392e-06, + "loss": 0.4015625, + "step": 590, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8659462140807143 + }, + { + "epoch": 0.2411591853277941, + "grad_norm": 0.44321754574775696, + "learning_rate": 8.879456848456145e-06, + "loss": 0.3928466796875, + "step": 595, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8695430999720835 + }, + { + "epoch": 0.24318573310365793, + "grad_norm": 0.3784260153770447, + "learning_rate": 8.85866796890021e-06, + "loss": 0.370458984375, + "step": 600, + "tIoU": 0.999885082244873, + "tRecall_30": 0.9998974800109863, + "tRecall_50": 0.9998974800109863, + "tRecall_70": 0.9998974800109863, + "token_acc": 0.8785528921249093 + }, + { + "epoch": 0.24521228087952174, + "grad_norm": 0.45076823234558105, + "learning_rate": 8.837712828413535e-06, + "loss": 0.3765380859375, + "step": 605, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8766975392882921 + }, + { + "epoch": 0.24723882865538555, + "grad_norm": 0.48391059041023254, + "learning_rate": 8.81659232990385e-06, + "loss": 0.3655517578125, + "step": 610, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8825125394927084 + }, + { + "epoch": 0.24926537643124938, + "grad_norm": 0.9391803741455078, + "learning_rate": 8.795307383403785e-06, + "loss": 0.375390625, + "step": 615, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8749314688760895 + }, + { + "epoch": 0.25129192420711316, + "grad_norm": 0.5365148782730103, + "learning_rate": 8.773858906031639e-06, + "loss": 0.402587890625, + "step": 620, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8673238029359772 + }, + { + "epoch": 0.253318471982977, + "grad_norm": 0.4107930660247803, + "learning_rate": 8.752247821951876e-06, + "loss": 0.3646484375, + "step": 625, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8803534683936145 + }, + { + "epoch": 0.25534501975884083, + "grad_norm": 0.36408987641334534, + "learning_rate": 8.730475062335298e-06, + "loss": 0.403076171875, + "step": 630, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8675791889438111 + }, + { + "epoch": 0.25737156753470464, + "grad_norm": 0.4031390845775604, + "learning_rate": 8.70854156531893e-06, + "loss": 0.37197265625, + "step": 635, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8779134815335501 + }, + { + "epoch": 0.25939811531056844, + "grad_norm": 0.486400842666626, + "learning_rate": 8.68644827596559e-06, + "loss": 0.3824951171875, + "step": 640, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8733007408146467 + }, + { + "epoch": 0.26142466308643225, + "grad_norm": 0.44566434621810913, + "learning_rate": 8.664196146223172e-06, + "loss": 0.3892578125, + "step": 645, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8704924857165266 + }, + { + "epoch": 0.26345121086229606, + "grad_norm": 0.3780648112297058, + "learning_rate": 8.641786134883631e-06, + "loss": 0.382177734375, + "step": 650, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8741021634541154 + }, + { + "epoch": 0.2654777586381599, + "grad_norm": 0.44057774543762207, + "learning_rate": 8.619219207541669e-06, + "loss": 0.3720947265625, + "step": 655, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8778030320781686 + }, + { + "epoch": 0.26750430641402373, + "grad_norm": 0.4715423285961151, + "learning_rate": 8.596496336553128e-06, + "loss": 0.371533203125, + "step": 660, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8801489222275704 + }, + { + "epoch": 0.26953085418988754, + "grad_norm": 0.5181432962417603, + "learning_rate": 8.573618500993095e-06, + "loss": 0.3737548828125, + "step": 665, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8751064228049731 + }, + { + "epoch": 0.27155740196575134, + "grad_norm": 0.41301727294921875, + "learning_rate": 8.550586686613715e-06, + "loss": 0.365869140625, + "step": 670, + "tIoU": 0.9999523162841797, + "tRecall_30": 0.9999526977539063, + "tRecall_50": 0.9999526977539063, + "tRecall_70": 0.9999526977539063, + "token_acc": 0.878654232390975 + }, + { + "epoch": 0.27358394974161515, + "grad_norm": 0.473439484834671, + "learning_rate": 8.527401885801723e-06, + "loss": 0.3747802734375, + "step": 675, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8762945655602113 + }, + { + "epoch": 0.27561049751747896, + "grad_norm": 0.425710529088974, + "learning_rate": 8.504065097535674e-06, + "loss": 0.383251953125, + "step": 680, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8726958420525884 + }, + { + "epoch": 0.27763704529334277, + "grad_norm": 0.4751035273075104, + "learning_rate": 8.48057732734291e-06, + "loss": 0.404248046875, + "step": 685, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8653982002918089 + }, + { + "epoch": 0.27966359306920663, + "grad_norm": 0.44467607140541077, + "learning_rate": 8.456939587256225e-06, + "loss": 0.381005859375, + "step": 690, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8726581191337954 + }, + { + "epoch": 0.28169014084507044, + "grad_norm": 0.48777613043785095, + "learning_rate": 8.433152895770269e-06, + "loss": 0.39541015625, + "step": 695, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8681833845913769 + }, + { + "epoch": 0.28371668862093424, + "grad_norm": 0.4859802722930908, + "learning_rate": 8.409218277797648e-06, + "loss": 0.3770263671875, + "step": 700, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8738482211415408 + }, + { + "epoch": 0.28574323639679805, + "grad_norm": 0.3302393853664398, + "learning_rate": 8.385136764624782e-06, + "loss": 0.3587890625, + "step": 705, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8811119122569451 + }, + { + "epoch": 0.28776978417266186, + "grad_norm": 0.39953362941741943, + "learning_rate": 8.360909393867455e-06, + "loss": 0.3962158203125, + "step": 710, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8683007948043105 + }, + { + "epoch": 0.28979633194852567, + "grad_norm": 0.2610228955745697, + "learning_rate": 8.336537209426115e-06, + "loss": 0.37705078125, + "step": 715, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8759010268027891 + }, + { + "epoch": 0.29182287972438953, + "grad_norm": 0.5030771493911743, + "learning_rate": 8.312021261440883e-06, + "loss": 0.360986328125, + "step": 720, + "tIoU": 0.9996023178100586, + "tRecall_30": 0.9997917175292969, + "tRecall_50": 0.9997917175292969, + "tRecall_70": 0.9997917175292969, + "token_acc": 0.8792668163208196 + }, + { + "epoch": 0.29384942750025334, + "grad_norm": 0.4199218153953552, + "learning_rate": 8.287362606246325e-06, + "loss": 0.3444091796875, + "step": 725, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8850844306174159 + }, + { + "epoch": 0.29587597527611714, + "grad_norm": 0.43417927622795105, + "learning_rate": 8.262562306325913e-06, + "loss": 0.380419921875, + "step": 730, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8734411766591651 + }, + { + "epoch": 0.29790252305198095, + "grad_norm": 0.40655168890953064, + "learning_rate": 8.237621430266267e-06, + "loss": 0.356982421875, + "step": 735, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8851337543171275 + }, + { + "epoch": 0.29992907082784476, + "grad_norm": 0.35124924778938293, + "learning_rate": 8.212541052711098e-06, + "loss": 0.351123046875, + "step": 740, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8840490141313778 + }, + { + "epoch": 0.30195561860370856, + "grad_norm": 0.37331146001815796, + "learning_rate": 8.187322254314907e-06, + "loss": 0.4010498046875, + "step": 745, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8676304091337044 + }, + { + "epoch": 0.30398216637957237, + "grad_norm": 0.536299467086792, + "learning_rate": 8.16196612169643e-06, + "loss": 0.3976318359375, + "step": 750, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8687691912291855 + }, + { + "epoch": 0.30600871415543623, + "grad_norm": 0.7118610143661499, + "learning_rate": 8.136473747391802e-06, + "loss": 0.37802734375, + "step": 755, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8742820838627701 + }, + { + "epoch": 0.30803526193130004, + "grad_norm": 0.4947682023048401, + "learning_rate": 8.110846229807497e-06, + "loss": 0.3813232421875, + "step": 760, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8758169106041493 + }, + { + "epoch": 0.31006180970716385, + "grad_norm": 0.36296963691711426, + "learning_rate": 8.085084673173e-06, + "loss": 0.384326171875, + "step": 765, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8708688664817646 + }, + { + "epoch": 0.31208835748302766, + "grad_norm": 0.4178961217403412, + "learning_rate": 8.059190187493216e-06, + "loss": 0.3591796875, + "step": 770, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8794477029858403 + }, + { + "epoch": 0.31411490525889146, + "grad_norm": 0.47259947657585144, + "learning_rate": 8.033163888500653e-06, + "loss": 0.3423828125, + "step": 775, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.884736740554099 + }, + { + "epoch": 0.31614145303475527, + "grad_norm": 0.36307206749916077, + "learning_rate": 8.007006897607346e-06, + "loss": 0.3815673828125, + "step": 780, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8737486629549381 + }, + { + "epoch": 0.31816800081061913, + "grad_norm": 0.47526979446411133, + "learning_rate": 7.980720341856532e-06, + "loss": 0.370849609375, + "step": 785, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8770150182461645 + }, + { + "epoch": 0.32019454858648294, + "grad_norm": 0.43980148434638977, + "learning_rate": 7.9543053538741e-06, + "loss": 0.3899658203125, + "step": 790, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8706937049796429 + }, + { + "epoch": 0.32222109636234675, + "grad_norm": 0.4804976284503937, + "learning_rate": 7.927763071819779e-06, + "loss": 0.336474609375, + "step": 795, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8857131373911246 + }, + { + "epoch": 0.32424764413821056, + "grad_norm": 0.5127283334732056, + "learning_rate": 7.901094639338098e-06, + "loss": 0.4080322265625, + "step": 800, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8650576159254072 + }, + { + "epoch": 0.32627419191407436, + "grad_norm": 0.5694597363471985, + "learning_rate": 7.874301205509112e-06, + "loss": 0.3904296875, + "step": 805, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8694054648160928 + }, + { + "epoch": 0.32830073968993817, + "grad_norm": 0.43124932050704956, + "learning_rate": 7.847383924798892e-06, + "loss": 0.3904541015625, + "step": 810, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.871555706607253 + }, + { + "epoch": 0.33032728746580203, + "grad_norm": 0.4303266406059265, + "learning_rate": 7.82034395700978e-06, + "loss": 0.390966796875, + "step": 815, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8692795917031997 + }, + { + "epoch": 0.33235383524166584, + "grad_norm": 0.3481966257095337, + "learning_rate": 7.793182467230415e-06, + "loss": 0.3556884765625, + "step": 820, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8809747012649367 + }, + { + "epoch": 0.33438038301752965, + "grad_norm": 0.5233892202377319, + "learning_rate": 7.765900625785532e-06, + "loss": 0.3641357421875, + "step": 825, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8789725840008653 + }, + { + "epoch": 0.33640693079339345, + "grad_norm": 0.4372534453868866, + "learning_rate": 7.738499608185536e-06, + "loss": 0.3589599609375, + "step": 830, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8846019057790832 + }, + { + "epoch": 0.33843347856925726, + "grad_norm": 0.42524704337120056, + "learning_rate": 7.710980595075855e-06, + "loss": 0.376806640625, + "step": 835, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8755772871051479 + }, + { + "epoch": 0.34046002634512107, + "grad_norm": 0.28774598240852356, + "learning_rate": 7.683344772186063e-06, + "loss": 0.33037109375, + "step": 840, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.892224320921029 + }, + { + "epoch": 0.3424865741209849, + "grad_norm": 0.4768100380897522, + "learning_rate": 7.655593330278798e-06, + "loss": 0.3840087890625, + "step": 845, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8781859667120646 + }, + { + "epoch": 0.34451312189684874, + "grad_norm": 0.37228870391845703, + "learning_rate": 7.627727465098442e-06, + "loss": 0.360986328125, + "step": 850, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8804808702639224 + }, + { + "epoch": 0.34653966967271255, + "grad_norm": 0.42258158326148987, + "learning_rate": 7.599748377319611e-06, + "loss": 0.383642578125, + "step": 855, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8718607702866956 + }, + { + "epoch": 0.34856621744857635, + "grad_norm": 0.5075936913490295, + "learning_rate": 7.5716572724954184e-06, + "loss": 0.403564453125, + "step": 860, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8665800600833172 + }, + { + "epoch": 0.35059276522444016, + "grad_norm": 0.3515145182609558, + "learning_rate": 7.543455361005526e-06, + "loss": 0.3584716796875, + "step": 865, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8810606243803352 + }, + { + "epoch": 0.35261931300030397, + "grad_norm": 0.4498981833457947, + "learning_rate": 7.5151438580039945e-06, + "loss": 0.3622802734375, + "step": 870, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8861151058022481 + }, + { + "epoch": 0.3546458607761678, + "grad_norm": 0.4505605697631836, + "learning_rate": 7.486723983366925e-06, + "loss": 0.3845703125, + "step": 875, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8717318269256821 + }, + { + "epoch": 0.35667240855203164, + "grad_norm": 0.6111229062080383, + "learning_rate": 7.4581969616399e-06, + "loss": 0.380859375, + "step": 880, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8744601628693373 + }, + { + "epoch": 0.35869895632789545, + "grad_norm": 0.5495724678039551, + "learning_rate": 7.4295640219852135e-06, + "loss": 0.4007568359375, + "step": 885, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8689283873573946 + }, + { + "epoch": 0.36072550410375925, + "grad_norm": 0.3697036802768707, + "learning_rate": 7.40082639812892e-06, + "loss": 0.3607421875, + "step": 890, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8797935193453291 + }, + { + "epoch": 0.36275205187962306, + "grad_norm": 0.5050045847892761, + "learning_rate": 7.371985328307664e-06, + "loss": 0.3994873046875, + "step": 895, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8684937988743339 + }, + { + "epoch": 0.36477859965548687, + "grad_norm": 0.42693156003952026, + "learning_rate": 7.343042055215342e-06, + "loss": 0.381591796875, + "step": 900, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8741249800223749 + }, + { + "epoch": 0.3668051474313507, + "grad_norm": 0.4217187762260437, + "learning_rate": 7.313997825949536e-06, + "loss": 0.37421875, + "step": 905, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8787105693100414 + }, + { + "epoch": 0.3688316952072145, + "grad_norm": 0.41110745072364807, + "learning_rate": 7.284853891957806e-06, + "loss": 0.3683837890625, + "step": 910, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8784763482289679 + }, + { + "epoch": 0.37085824298307835, + "grad_norm": 0.43751952052116394, + "learning_rate": 7.2556115089837435e-06, + "loss": 0.3561767578125, + "step": 915, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8809951857527238 + }, + { + "epoch": 0.37288479075894215, + "grad_norm": 0.3949474096298218, + "learning_rate": 7.226271937012883e-06, + "loss": 0.396240234375, + "step": 920, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8673576077580574 + }, + { + "epoch": 0.37491133853480596, + "grad_norm": 0.4348854124546051, + "learning_rate": 7.1968364402183985e-06, + "loss": 0.3632568359375, + "step": 925, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8787858087673649 + }, + { + "epoch": 0.37693788631066977, + "grad_norm": 0.45749151706695557, + "learning_rate": 7.167306286906645e-06, + "loss": 0.3761962890625, + "step": 930, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8756596442196003 + }, + { + "epoch": 0.3789644340865336, + "grad_norm": 0.4483352601528168, + "learning_rate": 7.137682749462496e-06, + "loss": 0.394189453125, + "step": 935, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8667820435842193 + }, + { + "epoch": 0.3809909818623974, + "grad_norm": 0.5248214602470398, + "learning_rate": 7.107967104294538e-06, + "loss": 0.3496337890625, + "step": 940, + "tIoU": 0.99993896484375, + "tRecall_30": 0.9999264717102051, + "tRecall_50": 0.9999264717102051, + "tRecall_70": 0.9999264717102051, + "token_acc": 0.8841316967787876 + }, + { + "epoch": 0.38301752963826124, + "grad_norm": 0.39933234453201294, + "learning_rate": 7.078160631780054e-06, + "loss": 0.37080078125, + "step": 945, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8746478070647529 + }, + { + "epoch": 0.38504407741412505, + "grad_norm": 0.49557754397392273, + "learning_rate": 7.048264616209869e-06, + "loss": 0.365966796875, + "step": 950, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8790204772034137 + }, + { + "epoch": 0.38707062518998886, + "grad_norm": 0.4162745773792267, + "learning_rate": 7.018280345733004e-06, + "loss": 0.3338623046875, + "step": 955, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.888543501924566 + }, + { + "epoch": 0.38909717296585267, + "grad_norm": 0.3654502332210541, + "learning_rate": 6.988209112301181e-06, + "loss": 0.338916015625, + "step": 960, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8869780634043012 + }, + { + "epoch": 0.3911237207417165, + "grad_norm": 0.3963611125946045, + "learning_rate": 6.958052211613147e-06, + "loss": 0.3736083984375, + "step": 965, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8755711663818746 + }, + { + "epoch": 0.3931502685175803, + "grad_norm": 0.5171037912368774, + "learning_rate": 6.9278109430588505e-06, + "loss": 0.3814208984375, + "step": 970, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8742617061438227 + }, + { + "epoch": 0.39517681629344414, + "grad_norm": 0.5404708385467529, + "learning_rate": 6.897486609663455e-06, + "loss": 0.365283203125, + "step": 975, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.882941194328911 + }, + { + "epoch": 0.39720336406930795, + "grad_norm": 0.527686357498169, + "learning_rate": 6.867080518031194e-06, + "loss": 0.39111328125, + "step": 980, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8707319961822223 + }, + { + "epoch": 0.39922991184517176, + "grad_norm": 0.5123360753059387, + "learning_rate": 6.836593978289064e-06, + "loss": 0.3810302734375, + "step": 985, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8738838872216664 + }, + { + "epoch": 0.40125645962103557, + "grad_norm": 0.37030547857284546, + "learning_rate": 6.80602830403039e-06, + "loss": 0.3386962890625, + "step": 990, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8885187714839158 + }, + { + "epoch": 0.4032830073968994, + "grad_norm": 0.5129815936088562, + "learning_rate": 6.775384812258212e-06, + "loss": 0.3621337890625, + "step": 995, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8785935775476954 + }, + { + "epoch": 0.4053095551727632, + "grad_norm": 0.48550745844841003, + "learning_rate": 6.744664823328545e-06, + "loss": 0.3682861328125, + "step": 1000, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8782230637497287 + }, + { + "epoch": 0.407336102948627, + "grad_norm": 0.360082745552063, + "learning_rate": 6.713869660893486e-06, + "loss": 0.3373046875, + "step": 1005, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8856287287482901 + }, + { + "epoch": 0.40936265072449085, + "grad_norm": 0.37603187561035156, + "learning_rate": 6.683000651844186e-06, + "loss": 0.3563232421875, + "step": 1010, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.880387902345518 + }, + { + "epoch": 0.41138919850035466, + "grad_norm": 0.3709280490875244, + "learning_rate": 6.652059126253665e-06, + "loss": 0.3497314453125, + "step": 1015, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.886997627845445 + }, + { + "epoch": 0.41341574627621847, + "grad_norm": 0.4052375555038452, + "learning_rate": 6.621046417319518e-06, + "loss": 0.381396484375, + "step": 1020, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8730041806398253 + }, + { + "epoch": 0.4154422940520823, + "grad_norm": 0.39146503806114197, + "learning_rate": 6.589963861306456e-06, + "loss": 0.3515869140625, + "step": 1025, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8837771654074887 + }, + { + "epoch": 0.4174688418279461, + "grad_norm": 0.40089619159698486, + "learning_rate": 6.558812797488745e-06, + "loss": 0.3480712890625, + "step": 1030, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8839427066060109 + }, + { + "epoch": 0.4194953896038099, + "grad_norm": 0.4794605076313019, + "learning_rate": 6.527594568092482e-06, + "loss": 0.3528564453125, + "step": 1035, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8843052146682948 + }, + { + "epoch": 0.42152193737967375, + "grad_norm": 0.5362907648086548, + "learning_rate": 6.4963105182377784e-06, + "loss": 0.33486328125, + "step": 1040, + "tIoU": 0.9999542236328125, + "tRecall_30": 0.9999536514282227, + "tRecall_50": 0.9999536514282227, + "tRecall_70": 0.9999536514282227, + "token_acc": 0.8922635725746582 + }, + { + "epoch": 0.42354848515553756, + "grad_norm": 0.35571110248565674, + "learning_rate": 6.464961995880788e-06, + "loss": 0.37939453125, + "step": 1045, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.877855904146303 + }, + { + "epoch": 0.42557503293140136, + "grad_norm": 0.36828193068504333, + "learning_rate": 6.433550351755639e-06, + "loss": 0.3494384765625, + "step": 1050, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8832511866683361 + }, + { + "epoch": 0.42760158070726517, + "grad_norm": 0.35322102904319763, + "learning_rate": 6.402076939316223e-06, + "loss": 0.3216064453125, + "step": 1055, + "tIoU": 0.99375, + "tRecall_30": 0.99375, + "tRecall_50": 0.99375, + "tRecall_70": 0.99375, + "token_acc": 0.8966078344163241 + }, + { + "epoch": 0.429628128483129, + "grad_norm": 0.5031628608703613, + "learning_rate": 6.370543114677887e-06, + "loss": 0.3531982421875, + "step": 1060, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8816162500961772 + }, + { + "epoch": 0.4316546762589928, + "grad_norm": 0.409626305103302, + "learning_rate": 6.338950236558996e-06, + "loss": 0.3392578125, + "step": 1065, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8864688797464766 + }, + { + "epoch": 0.43368122403485665, + "grad_norm": 0.41023844480514526, + "learning_rate": 6.307299666222391e-06, + "loss": 0.3597412109375, + "step": 1070, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8799794250442933 + }, + { + "epoch": 0.43570777181072046, + "grad_norm": 0.4652061462402344, + "learning_rate": 6.275592767416734e-06, + "loss": 0.3640380859375, + "step": 1075, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8776116953453434 + }, + { + "epoch": 0.43773431958658426, + "grad_norm": 0.4822935461997986, + "learning_rate": 6.243830906317749e-06, + "loss": 0.36923828125, + "step": 1080, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8774388126503158 + }, + { + "epoch": 0.43976086736244807, + "grad_norm": 0.4883968234062195, + "learning_rate": 6.212015451469357e-06, + "loss": 0.3608154296875, + "step": 1085, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8790962323901034 + }, + { + "epoch": 0.4417874151383119, + "grad_norm": 0.43007761240005493, + "learning_rate": 6.180147773724706e-06, + "loss": 0.3548828125, + "step": 1090, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8837881721101605 + }, + { + "epoch": 0.4438139629141757, + "grad_norm": 0.4888928234577179, + "learning_rate": 6.148229246187107e-06, + "loss": 0.3470703125, + "step": 1095, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.88543772516882 + }, + { + "epoch": 0.4458405106900395, + "grad_norm": 0.4666309058666229, + "learning_rate": 6.116261244150868e-06, + "loss": 0.396435546875, + "step": 1100, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8679060399031002 + }, + { + "epoch": 0.44786705846590336, + "grad_norm": 0.42913663387298584, + "learning_rate": 6.084245145042039e-06, + "loss": 0.353662109375, + "step": 1105, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8826747113821377 + }, + { + "epoch": 0.44989360624176716, + "grad_norm": 0.482410728931427, + "learning_rate": 6.052182328359056e-06, + "loss": 0.33896484375, + "step": 1110, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8857838859345668 + }, + { + "epoch": 0.45192015401763097, + "grad_norm": 0.4276754558086395, + "learning_rate": 6.0200741756133105e-06, + "loss": 0.3577880859375, + "step": 1115, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8800624976841274 + }, + { + "epoch": 0.4539467017934948, + "grad_norm": 0.4833751320838928, + "learning_rate": 5.987922070269612e-06, + "loss": 0.32001953125, + "step": 1120, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8940394473889999 + }, + { + "epoch": 0.4559732495693586, + "grad_norm": 0.4228677451610565, + "learning_rate": 5.95572739768659e-06, + "loss": 0.3525634765625, + "step": 1125, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8830779480346436 + }, + { + "epoch": 0.4579997973452224, + "grad_norm": 0.4276502728462219, + "learning_rate": 5.923491545056992e-06, + "loss": 0.3637939453125, + "step": 1130, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8791735003721488 + }, + { + "epoch": 0.46002634512108626, + "grad_norm": 0.46918392181396484, + "learning_rate": 5.891215901347922e-06, + "loss": 0.3595947265625, + "step": 1135, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8808175171468626 + }, + { + "epoch": 0.46205289289695006, + "grad_norm": 0.43709152936935425, + "learning_rate": 5.85890185724098e-06, + "loss": 0.3795166015625, + "step": 1140, + "tIoU": 0.99375, + "tRecall_30": 0.99375, + "tRecall_50": 0.99375, + "tRecall_70": 0.99375, + "token_acc": 0.872928022166402 + }, + { + "epoch": 0.46407944067281387, + "grad_norm": 0.4206150472164154, + "learning_rate": 5.826550805072358e-06, + "loss": 0.3873046875, + "step": 1145, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8712319938126389 + }, + { + "epoch": 0.4661059884486777, + "grad_norm": 0.3167666494846344, + "learning_rate": 5.794164138772832e-06, + "loss": 0.337109375, + "step": 1150, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8865754147337275 + }, + { + "epoch": 0.4681325362245415, + "grad_norm": 0.4532758891582489, + "learning_rate": 5.76174325380771e-06, + "loss": 0.36435546875, + "step": 1155, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8779640518250711 + }, + { + "epoch": 0.4701590840004053, + "grad_norm": 0.4313461184501648, + "learning_rate": 5.729289547116701e-06, + "loss": 0.389697265625, + "step": 1160, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8700403579575364 + }, + { + "epoch": 0.4721856317762691, + "grad_norm": 0.4814442992210388, + "learning_rate": 5.696804417053724e-06, + "loss": 0.3765380859375, + "step": 1165, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8757619796016335 + }, + { + "epoch": 0.47421217955213296, + "grad_norm": 0.3513696491718292, + "learning_rate": 5.664289263326659e-06, + "loss": 0.3399658203125, + "step": 1170, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8893098409021976 + }, + { + "epoch": 0.47623872732799677, + "grad_norm": 0.40614351630210876, + "learning_rate": 5.631745486937035e-06, + "loss": 0.3510986328125, + "step": 1175, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8833963519767688 + }, + { + "epoch": 0.4782652751038606, + "grad_norm": 0.5500504374504089, + "learning_rate": 5.599174490119661e-06, + "loss": 0.3646240234375, + "step": 1180, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8787946444647706 + }, + { + "epoch": 0.4802918228797244, + "grad_norm": 0.5975911021232605, + "learning_rate": 5.566577676282216e-06, + "loss": 0.3532470703125, + "step": 1185, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8857554384480847 + }, + { + "epoch": 0.4823183706555882, + "grad_norm": 0.3443179130554199, + "learning_rate": 5.533956449944771e-06, + "loss": 0.32177734375, + "step": 1190, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8915946582875098 + }, + { + "epoch": 0.484344918431452, + "grad_norm": 0.3399013578891754, + "learning_rate": 5.501312216679271e-06, + "loss": 0.346630859375, + "step": 1195, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8862332459567733 + }, + { + "epoch": 0.48637146620731586, + "grad_norm": 0.4984676241874695, + "learning_rate": 5.468646383048978e-06, + "loss": 0.3672119140625, + "step": 1200, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8770362908179562 + }, + { + "epoch": 0.48839801398317967, + "grad_norm": 0.3607572913169861, + "learning_rate": 5.435960356547866e-06, + "loss": 0.3676025390625, + "step": 1205, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.878624072395896 + }, + { + "epoch": 0.4904245617590435, + "grad_norm": 0.4026191234588623, + "learning_rate": 5.403255545539967e-06, + "loss": 0.3791259765625, + "step": 1210, + "tIoU": 0.9997499465942383, + "tRecall_30": 0.9995536804199219, + "tRecall_50": 0.9995536804199219, + "tRecall_70": 0.9995536804199219, + "token_acc": 0.874137520818463 + }, + { + "epoch": 0.4924511095349073, + "grad_norm": 0.4158448576927185, + "learning_rate": 5.370533359198697e-06, + "loss": 0.3332763671875, + "step": 1215, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8925485768613702 + }, + { + "epoch": 0.4944776573107711, + "grad_norm": 0.4042837917804718, + "learning_rate": 5.337795207446134e-06, + "loss": 0.3458740234375, + "step": 1220, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8848929158802503 + }, + { + "epoch": 0.4965042050866349, + "grad_norm": 0.38991838693618774, + "learning_rate": 5.305042500892265e-06, + "loss": 0.334619140625, + "step": 1225, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8871413548783782 + }, + { + "epoch": 0.49853075286249876, + "grad_norm": 0.38406676054000854, + "learning_rate": 5.272276650774211e-06, + "loss": 0.383642578125, + "step": 1230, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8712314225053078 + }, + { + "epoch": 0.5005573006383626, + "grad_norm": 0.4755161702632904, + "learning_rate": 5.23949906889542e-06, + "loss": 0.3812255859375, + "step": 1235, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8735439060691251 + }, + { + "epoch": 0.5025838484142263, + "grad_norm": 0.48450133204460144, + "learning_rate": 5.20671116756483e-06, + "loss": 0.3880126953125, + "step": 1240, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8706281504459092 + }, + { + "epoch": 0.5046103961900902, + "grad_norm": 0.4431735873222351, + "learning_rate": 5.173914359536026e-06, + "loss": 0.3583251953125, + "step": 1245, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8804128704576395 + }, + { + "epoch": 0.506636943965954, + "grad_norm": 0.4925408959388733, + "learning_rate": 5.141110057946357e-06, + "loss": 0.353564453125, + "step": 1250, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.881574081411628 + }, + { + "epoch": 0.5086634917418178, + "grad_norm": 0.39856627583503723, + "learning_rate": 5.1082996762560495e-06, + "loss": 0.36533203125, + "step": 1255, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8775912245354824 + }, + { + "epoch": 0.5106900395176817, + "grad_norm": 0.41317877173423767, + "learning_rate": 5.075484628187315e-06, + "loss": 0.35595703125, + "step": 1260, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8816695158533046 + }, + { + "epoch": 0.5127165872935454, + "grad_norm": 0.4024195969104767, + "learning_rate": 5.042666327663421e-06, + "loss": 0.3582763671875, + "step": 1265, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8803793129712849 + }, + { + "epoch": 0.5147431350694093, + "grad_norm": 0.513283371925354, + "learning_rate": 5.009846188747782e-06, + "loss": 0.3602783203125, + "step": 1270, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8798386381166255 + }, + { + "epoch": 0.516769682845273, + "grad_norm": 0.44748246669769287, + "learning_rate": 4.977025625583017e-06, + "loss": 0.376123046875, + "step": 1275, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8758046600018736 + }, + { + "epoch": 0.5187962306211369, + "grad_norm": 0.3967839777469635, + "learning_rate": 4.944206052330032e-06, + "loss": 0.35078125, + "step": 1280, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8834837609267288 + }, + { + "epoch": 0.5208227783970008, + "grad_norm": 0.3833644986152649, + "learning_rate": 4.911388883107076e-06, + "loss": 0.3705810546875, + "step": 1285, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8753450905744576 + }, + { + "epoch": 0.5228493261728645, + "grad_norm": 0.3436470031738281, + "learning_rate": 4.878575531928815e-06, + "loss": 0.3374267578125, + "step": 1290, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8864620406000278 + }, + { + "epoch": 0.5248758739487284, + "grad_norm": 0.6544244885444641, + "learning_rate": 4.845767412645405e-06, + "loss": 0.3248046875, + "step": 1295, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.891281923817809 + }, + { + "epoch": 0.5269024217245921, + "grad_norm": 0.378275066614151, + "learning_rate": 4.81296593888157e-06, + "loss": 0.3455322265625, + "step": 1300, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8847039987503905 + }, + { + "epoch": 0.528928969500456, + "grad_norm": 0.4009753167629242, + "learning_rate": 4.780172523975697e-06, + "loss": 0.3330810546875, + "step": 1305, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8884986977137642 + }, + { + "epoch": 0.5309555172763198, + "grad_norm": 0.46430066227912903, + "learning_rate": 4.747388580918934e-06, + "loss": 0.3642333984375, + "step": 1310, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8783965865708511 + }, + { + "epoch": 0.5329820650521836, + "grad_norm": 0.49786871671676636, + "learning_rate": 4.71461552229431e-06, + "loss": 0.3664794921875, + "step": 1315, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8775798506769107 + }, + { + "epoch": 0.5350086128280475, + "grad_norm": 0.46610528230667114, + "learning_rate": 4.68185476021587e-06, + "loss": 0.3385986328125, + "step": 1320, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8856598940033139 + }, + { + "epoch": 0.5370351606039112, + "grad_norm": 0.45692896842956543, + "learning_rate": 4.649107706267823e-06, + "loss": 0.3858642578125, + "step": 1325, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8724768936577074 + }, + { + "epoch": 0.5390617083797751, + "grad_norm": 0.5358167886734009, + "learning_rate": 4.616375771443743e-06, + "loss": 0.37216796875, + "step": 1330, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8778596081550957 + }, + { + "epoch": 0.5410882561556388, + "grad_norm": 0.4610866904258728, + "learning_rate": 4.583660366085735e-06, + "loss": 0.3567626953125, + "step": 1335, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8821897267269218 + }, + { + "epoch": 0.5431148039315027, + "grad_norm": 0.4700048863887787, + "learning_rate": 4.5509628998237085e-06, + "loss": 0.3708984375, + "step": 1340, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8747563915390623 + }, + { + "epoch": 0.5451413517073666, + "grad_norm": 0.4478241503238678, + "learning_rate": 4.518284781514603e-06, + "loss": 0.371826171875, + "step": 1345, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8773185057988198 + }, + { + "epoch": 0.5471678994832303, + "grad_norm": 0.4090299904346466, + "learning_rate": 4.485627419181714e-06, + "loss": 0.3501953125, + "step": 1350, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8831489211736755 + }, + { + "epoch": 0.5491944472590942, + "grad_norm": 0.4647216200828552, + "learning_rate": 4.452992219953999e-06, + "loss": 0.34912109375, + "step": 1355, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8830212417158579 + }, + { + "epoch": 0.5512209950349579, + "grad_norm": 0.5491527318954468, + "learning_rate": 4.42038059000547e-06, + "loss": 0.350537109375, + "step": 1360, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8814508260152754 + }, + { + "epoch": 0.5532475428108218, + "grad_norm": 0.3985271751880646, + "learning_rate": 4.387793934494583e-06, + "loss": 0.3564453125, + "step": 1365, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8810057091206359 + }, + { + "epoch": 0.5552740905866855, + "grad_norm": 0.38414013385772705, + "learning_rate": 4.355233657503717e-06, + "loss": 0.338720703125, + "step": 1370, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8854592141406903 + }, + { + "epoch": 0.5573006383625494, + "grad_norm": 0.467517614364624, + "learning_rate": 4.322701161978648e-06, + "loss": 0.373388671875, + "step": 1375, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8745456194576039 + }, + { + "epoch": 0.5593271861384133, + "grad_norm": 0.41305845975875854, + "learning_rate": 4.29019784966813e-06, + "loss": 0.3896728515625, + "step": 1380, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8690092846025869 + }, + { + "epoch": 0.561353733914277, + "grad_norm": 0.45166894793510437, + "learning_rate": 4.257725121063464e-06, + "loss": 0.3476318359375, + "step": 1385, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8838776329382966 + }, + { + "epoch": 0.5633802816901409, + "grad_norm": 0.3921132981777191, + "learning_rate": 4.225284375338187e-06, + "loss": 0.3684814453125, + "step": 1390, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8769061119557345 + }, + { + "epoch": 0.5654068294660046, + "grad_norm": 0.6856555938720703, + "learning_rate": 4.192877010287755e-06, + "loss": 0.3556640625, + "step": 1395, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8795452874700099 + }, + { + "epoch": 0.5674333772418685, + "grad_norm": 0.4415869414806366, + "learning_rate": 4.160504422269343e-06, + "loss": 0.3482666015625, + "step": 1400, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8845478325859492 + }, + { + "epoch": 0.5694599250177322, + "grad_norm": 0.44592997431755066, + "learning_rate": 4.128168006141651e-06, + "loss": 0.359033203125, + "step": 1405, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8789114899636594 + }, + { + "epoch": 0.5714864727935961, + "grad_norm": 0.420045405626297, + "learning_rate": 4.09586915520483e-06, + "loss": 0.3491943359375, + "step": 1410, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.882240717334796 + }, + { + "epoch": 0.57351302056946, + "grad_norm": 0.348161518573761, + "learning_rate": 4.063609261140425e-06, + "loss": 0.36396484375, + "step": 1415, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8796527000915285 + }, + { + "epoch": 0.5755395683453237, + "grad_norm": 0.5775684118270874, + "learning_rate": 4.031389713951432e-06, + "loss": 0.3662109375, + "step": 1420, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8783637725837949 + }, + { + "epoch": 0.5775661161211876, + "grad_norm": 0.39185187220573425, + "learning_rate": 3.999211901902382e-06, + "loss": 0.369384765625, + "step": 1425, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8774693489029958 + }, + { + "epoch": 0.5795926638970513, + "grad_norm": 0.49676889181137085, + "learning_rate": 3.967077211459547e-06, + "loss": 0.349609375, + "step": 1430, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8816592219550906 + }, + { + "epoch": 0.5816192116729152, + "grad_norm": 0.42753615975379944, + "learning_rate": 3.934987027231185e-06, + "loss": 0.360302734375, + "step": 1435, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8803608665251746 + }, + { + "epoch": 0.5836457594487791, + "grad_norm": 0.41853469610214233, + "learning_rate": 3.90294273190789e-06, + "loss": 0.3525634765625, + "step": 1440, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8811438899178998 + }, + { + "epoch": 0.5856723072246428, + "grad_norm": 0.3808477818965912, + "learning_rate": 3.870945706203008e-06, + "loss": 0.3712646484375, + "step": 1445, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8750312781503353 + }, + { + "epoch": 0.5876988550005067, + "grad_norm": 0.3081310987472534, + "learning_rate": 3.8389973287931486e-06, + "loss": 0.3540283203125, + "step": 1450, + "tIoU": 0.9997791290283203, + "tRecall_30": 0.9997023582458496, + "tRecall_50": 0.9997023582458496, + "tRecall_70": 0.9997023582458496, + "token_acc": 0.8833477455219271 + }, + { + "epoch": 0.5897254027763704, + "grad_norm": 0.3728139102458954, + "learning_rate": 3.8070989762587805e-06, + "loss": 0.3547607421875, + "step": 1455, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8799670478774957 + }, + { + "epoch": 0.5917519505522343, + "grad_norm": 0.5454410910606384, + "learning_rate": 3.7752520230249216e-06, + "loss": 0.354736328125, + "step": 1460, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8818026764628706 + }, + { + "epoch": 0.593778498328098, + "grad_norm": 0.3359200656414032, + "learning_rate": 3.743457841301909e-06, + "loss": 0.3484619140625, + "step": 1465, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.881686928154548 + }, + { + "epoch": 0.5958050461039619, + "grad_norm": 0.4866096079349518, + "learning_rate": 3.7117178010262843e-06, + "loss": 0.36337890625, + "step": 1470, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8767980029235277 + }, + { + "epoch": 0.5978315938798258, + "grad_norm": 0.42000898718833923, + "learning_rate": 3.680033269801759e-06, + "loss": 0.3525390625, + "step": 1475, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8828758953386568 + }, + { + "epoch": 0.5998581416556895, + "grad_norm": 0.4449663460254669, + "learning_rate": 3.648405612840291e-06, + "loss": 0.317041015625, + "step": 1480, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8941719450487374 + }, + { + "epoch": 0.6018846894315534, + "grad_norm": 0.4433095157146454, + "learning_rate": 3.6168361929032584e-06, + "loss": 0.3648193359375, + "step": 1485, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8779625300451088 + }, + { + "epoch": 0.6039112372074171, + "grad_norm": 0.31316468119621277, + "learning_rate": 3.5853263702427465e-06, + "loss": 0.336474609375, + "step": 1490, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8874035573171796 + }, + { + "epoch": 0.605937784983281, + "grad_norm": 0.3436121940612793, + "learning_rate": 3.5538775025429285e-06, + "loss": 0.34521484375, + "step": 1495, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8830729953564668 + }, + { + "epoch": 0.6079643327591447, + "grad_norm": 0.3756560981273651, + "learning_rate": 3.522490944861578e-06, + "loss": 0.3583984375, + "step": 1500, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8821701099263564 + }, + { + "epoch": 0.6099908805350086, + "grad_norm": 0.43999379873275757, + "learning_rate": 3.4911680495716703e-06, + "loss": 0.3664794921875, + "step": 1505, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8774288485334939 + }, + { + "epoch": 0.6120174283108725, + "grad_norm": 0.48006972670555115, + "learning_rate": 3.459910166303123e-06, + "loss": 0.3483642578125, + "step": 1510, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8825045995932991 + }, + { + "epoch": 0.6140439760867362, + "grad_norm": 0.37062063813209534, + "learning_rate": 3.4287186418846353e-06, + "loss": 0.3367919921875, + "step": 1515, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8869997197058701 + }, + { + "epoch": 0.6160705238626001, + "grad_norm": 0.45247992873191833, + "learning_rate": 3.397594820285661e-06, + "loss": 0.3341064453125, + "step": 1520, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8887578872530142 + }, + { + "epoch": 0.6180970716384638, + "grad_norm": 0.3768942654132843, + "learning_rate": 3.366540042558497e-06, + "loss": 0.3563232421875, + "step": 1525, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8823201159673136 + }, + { + "epoch": 0.6201236194143277, + "grad_norm": 0.4086240828037262, + "learning_rate": 3.3355556467805063e-06, + "loss": 0.350146484375, + "step": 1530, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8824231340428154 + }, + { + "epoch": 0.6221501671901916, + "grad_norm": 0.39199697971343994, + "learning_rate": 3.304642967996452e-06, + "loss": 0.3452880859375, + "step": 1535, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8839277431803143 + }, + { + "epoch": 0.6241767149660553, + "grad_norm": 0.44045987725257874, + "learning_rate": 3.2738033381609887e-06, + "loss": 0.3403564453125, + "step": 1540, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.885882583705324 + }, + { + "epoch": 0.6262032627419192, + "grad_norm": 0.3382072448730469, + "learning_rate": 3.243038086081256e-06, + "loss": 0.3364990234375, + "step": 1545, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8862242631575911 + }, + { + "epoch": 0.6282298105177829, + "grad_norm": 0.33277419209480286, + "learning_rate": 3.2123485373596387e-06, + "loss": 0.3208984375, + "step": 1550, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8928338244895798 + }, + { + "epoch": 0.6302563582936468, + "grad_norm": 0.5623494982719421, + "learning_rate": 3.181736014336633e-06, + "loss": 0.3424072265625, + "step": 1555, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8854034061041338 + }, + { + "epoch": 0.6322829060695105, + "grad_norm": 0.40705209970474243, + "learning_rate": 3.151201836033887e-06, + "loss": 0.3233154296875, + "step": 1560, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8919693124471282 + }, + { + "epoch": 0.6343094538453744, + "grad_norm": 0.3886021077632904, + "learning_rate": 3.1207473180973537e-06, + "loss": 0.32705078125, + "step": 1565, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8905853670549413 + }, + { + "epoch": 0.6363360016212383, + "grad_norm": 0.45216310024261475, + "learning_rate": 3.0903737727406103e-06, + "loss": 0.340234375, + "step": 1570, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8851567758256551 + }, + { + "epoch": 0.638362549397102, + "grad_norm": 0.42403921484947205, + "learning_rate": 3.0600825086883157e-06, + "loss": 0.345654296875, + "step": 1575, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8841203080129101 + }, + { + "epoch": 0.6403890971729659, + "grad_norm": 0.43884193897247314, + "learning_rate": 3.029874831119821e-06, + "loss": 0.3524658203125, + "step": 1580, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8821244395373574 + }, + { + "epoch": 0.6424156449488296, + "grad_norm": 0.38225412368774414, + "learning_rate": 2.9997520416129317e-06, + "loss": 0.3519287109375, + "step": 1585, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.881298560950009 + }, + { + "epoch": 0.6444421927246935, + "grad_norm": 0.4440082311630249, + "learning_rate": 2.9697154380878267e-06, + "loss": 0.3475830078125, + "step": 1590, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8843408659299762 + }, + { + "epoch": 0.6464687405005572, + "grad_norm": 0.38897669315338135, + "learning_rate": 2.9397663147511333e-06, + "loss": 0.3405517578125, + "step": 1595, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8871316466216582 + }, + { + "epoch": 0.6484952882764211, + "grad_norm": 0.4603720009326935, + "learning_rate": 2.9099059620401647e-06, + "loss": 0.3591064453125, + "step": 1600, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8784650512461998 + }, + { + "epoch": 0.650521836052285, + "grad_norm": 0.4842081665992737, + "learning_rate": 2.88013566656731e-06, + "loss": 0.3609130859375, + "step": 1605, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8791378525786873 + }, + { + "epoch": 0.6525483838281487, + "grad_norm": 0.3952023983001709, + "learning_rate": 2.8504567110646166e-06, + "loss": 0.3239501953125, + "step": 1610, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8904520420960794 + }, + { + "epoch": 0.6545749316040126, + "grad_norm": 0.4561847746372223, + "learning_rate": 2.8208703743284928e-06, + "loss": 0.3415283203125, + "step": 1615, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8860618126198743 + }, + { + "epoch": 0.6566014793798763, + "grad_norm": 0.47599977254867554, + "learning_rate": 2.7913779311646337e-06, + "loss": 0.3582275390625, + "step": 1620, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8794347726568916 + }, + { + "epoch": 0.6586280271557402, + "grad_norm": 0.4711098372936249, + "learning_rate": 2.761980652333073e-06, + "loss": 0.320068359375, + "step": 1625, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8925188697393882 + }, + { + "epoch": 0.6606545749316041, + "grad_norm": 0.47064128518104553, + "learning_rate": 2.73267980449344e-06, + "loss": 0.324853515625, + "step": 1630, + "tIoU": 0.9998732566833496, + "tRecall_30": 0.9995833396911621, + "tRecall_50": 0.9995833396911621, + "tRecall_70": 0.9995833396911621, + "token_acc": 0.8897695675796606 + }, + { + "epoch": 0.6626811227074678, + "grad_norm": 0.47430020570755005, + "learning_rate": 2.70347665015038e-06, + "loss": 0.3352783203125, + "step": 1635, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8870509525156498 + }, + { + "epoch": 0.6647076704833317, + "grad_norm": 0.42994794249534607, + "learning_rate": 2.6743724475991584e-06, + "loss": 0.3516845703125, + "step": 1640, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.881572338841925 + }, + { + "epoch": 0.6667342182591954, + "grad_norm": 0.37945184111595154, + "learning_rate": 2.6453684508714307e-06, + "loss": 0.3171630859375, + "step": 1645, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8962419161172084 + }, + { + "epoch": 0.6687607660350593, + "grad_norm": 0.4700915515422821, + "learning_rate": 2.6164659096812285e-06, + "loss": 0.3476806640625, + "step": 1650, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.88284849865816 + }, + { + "epoch": 0.670787313810923, + "grad_norm": 0.3228314518928528, + "learning_rate": 2.5876660693710994e-06, + "loss": 0.323876953125, + "step": 1655, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8916923522350968 + }, + { + "epoch": 0.6728138615867869, + "grad_norm": 0.4141518771648407, + "learning_rate": 2.5589701708584526e-06, + "loss": 0.3321533203125, + "step": 1660, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8889392826343863 + }, + { + "epoch": 0.6748404093626508, + "grad_norm": 0.5032157897949219, + "learning_rate": 2.5303794505820863e-06, + "loss": 0.3514404296875, + "step": 1665, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8822658696270993 + }, + { + "epoch": 0.6768669571385145, + "grad_norm": 0.5460382699966431, + "learning_rate": 2.501895140448919e-06, + "loss": 0.3413818359375, + "step": 1670, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8864838250847952 + }, + { + "epoch": 0.6788935049143784, + "grad_norm": 0.4051891565322876, + "learning_rate": 2.4735184677809054e-06, + "loss": 0.323388671875, + "step": 1675, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8904025200370893 + }, + { + "epoch": 0.6809200526902421, + "grad_norm": 0.4227135479450226, + "learning_rate": 2.4452506552621575e-06, + "loss": 0.3535400390625, + "step": 1680, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8815619943211309 + }, + { + "epoch": 0.682946600466106, + "grad_norm": 0.4865545630455017, + "learning_rate": 2.4170929208862525e-06, + "loss": 0.34765625, + "step": 1685, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8828536786093468 + }, + { + "epoch": 0.6849731482419698, + "grad_norm": 0.4662053883075714, + "learning_rate": 2.3890464779037658e-06, + "loss": 0.3500244140625, + "step": 1690, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8823926432126887 + }, + { + "epoch": 0.6869996960178336, + "grad_norm": 0.5017711520195007, + "learning_rate": 2.361112534769986e-06, + "loss": 0.339892578125, + "step": 1695, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8856021181411485 + }, + { + "epoch": 0.6890262437936975, + "grad_norm": 0.42811140418052673, + "learning_rate": 2.3332922950928495e-06, + "loss": 0.354833984375, + "step": 1700, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8817200279273888 + }, + { + "epoch": 0.6910527915695612, + "grad_norm": 0.45597004890441895, + "learning_rate": 2.305586957581073e-06, + "loss": 0.3400634765625, + "step": 1705, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8865903430650272 + }, + { + "epoch": 0.6930793393454251, + "grad_norm": 0.4592193365097046, + "learning_rate": 2.2779977159925137e-06, + "loss": 0.32783203125, + "step": 1710, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8893759222289732 + }, + { + "epoch": 0.6951058871212888, + "grad_norm": 0.5005931258201599, + "learning_rate": 2.250525759082728e-06, + "loss": 0.3681396484375, + "step": 1715, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8778964980174249 + }, + { + "epoch": 0.6971324348971527, + "grad_norm": 0.2610883414745331, + "learning_rate": 2.2231722705537513e-06, + "loss": 0.292578125, + "step": 1720, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.9055491418456092 + }, + { + "epoch": 0.6991589826730166, + "grad_norm": 0.4703691601753235, + "learning_rate": 2.1959384290030912e-06, + "loss": 0.35205078125, + "step": 1725, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8827783713387997 + }, + { + "epoch": 0.7011855304488803, + "grad_norm": 0.3500620126724243, + "learning_rate": 2.1688254078729524e-06, + "loss": 0.32626953125, + "step": 1730, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8898699068503675 + }, + { + "epoch": 0.7032120782247442, + "grad_norm": 0.3912479877471924, + "learning_rate": 2.141834375399671e-06, + "loss": 0.358837890625, + "step": 1735, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8790342914775593 + }, + { + "epoch": 0.7052386260006079, + "grad_norm": 0.4508066475391388, + "learning_rate": 2.1149664945633797e-06, + "loss": 0.3556884765625, + "step": 1740, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.880684688811495 + }, + { + "epoch": 0.7072651737764718, + "grad_norm": 0.3865109384059906, + "learning_rate": 2.0882229230378916e-06, + "loss": 0.3451904296875, + "step": 1745, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8843628924774509 + }, + { + "epoch": 0.7092917215523356, + "grad_norm": 0.46912604570388794, + "learning_rate": 2.0616048131408297e-06, + "loss": 0.302392578125, + "step": 1750, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.900628045174191 + }, + { + "epoch": 0.7113182693281994, + "grad_norm": 0.4662412703037262, + "learning_rate": 2.035113311783967e-06, + "loss": 0.34765625, + "step": 1755, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.882717010509926 + }, + { + "epoch": 0.7133448171040633, + "grad_norm": 0.4677916467189789, + "learning_rate": 2.0087495604238165e-06, + "loss": 0.3470947265625, + "step": 1760, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8837272721507348 + }, + { + "epoch": 0.715371364879927, + "grad_norm": 0.5265308618545532, + "learning_rate": 1.9825146950124367e-06, + "loss": 0.3470703125, + "step": 1765, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.882448854961832 + }, + { + "epoch": 0.7173979126557909, + "grad_norm": 0.5809183716773987, + "learning_rate": 1.9564098459485003e-06, + "loss": 0.3476318359375, + "step": 1770, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8854824391402744 + }, + { + "epoch": 0.7194244604316546, + "grad_norm": 0.44178304076194763, + "learning_rate": 1.930436138028578e-06, + "loss": 0.341748046875, + "step": 1775, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8857181884786847 + }, + { + "epoch": 0.7214510082075185, + "grad_norm": 0.3675595819950104, + "learning_rate": 1.9045946903986801e-06, + "loss": 0.3384765625, + "step": 1780, + "tIoU": 0.9989814758300781, + "tRecall_30": 0.9996322631835938, + "tRecall_50": 0.9996322631835938, + "tRecall_70": 0.9996322631835938, + "token_acc": 0.8861089376698108 + }, + { + "epoch": 0.7234775559833823, + "grad_norm": 0.47204509377479553, + "learning_rate": 1.878886616506026e-06, + "loss": 0.3569091796875, + "step": 1785, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8793076633963633 + }, + { + "epoch": 0.7255041037592461, + "grad_norm": 0.3641361892223358, + "learning_rate": 1.85331302405108e-06, + "loss": 0.3317626953125, + "step": 1790, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8881008958559753 + }, + { + "epoch": 0.72753065153511, + "grad_norm": 0.4817829728126526, + "learning_rate": 1.827875014939815e-06, + "loss": 0.3068359375, + "step": 1795, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8976465497883155 + }, + { + "epoch": 0.7295571993109737, + "grad_norm": 0.4408557116985321, + "learning_rate": 1.802573685236239e-06, + "loss": 0.3442626953125, + "step": 1800, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8833064053786418 + }, + { + "epoch": 0.7315837470868376, + "grad_norm": 0.4439926743507385, + "learning_rate": 1.77741012511516e-06, + "loss": 0.311279296875, + "step": 1805, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8974859360066844 + }, + { + "epoch": 0.7336102948627014, + "grad_norm": 0.4358379542827606, + "learning_rate": 1.7523854188152235e-06, + "loss": 0.3204345703125, + "step": 1810, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8913284952606635 + }, + { + "epoch": 0.7356368426385652, + "grad_norm": 0.36242666840553284, + "learning_rate": 1.7275006445921889e-06, + "loss": 0.3365966796875, + "step": 1815, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8881554550174293 + }, + { + "epoch": 0.737663390414429, + "grad_norm": 0.4304690659046173, + "learning_rate": 1.702756874672473e-06, + "loss": 0.3471435546875, + "step": 1820, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8829572785903128 + }, + { + "epoch": 0.7396899381902928, + "grad_norm": 0.40860116481781006, + "learning_rate": 1.6781551752069441e-06, + "loss": 0.3568359375, + "step": 1825, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8813770378742888 + }, + { + "epoch": 0.7417164859661567, + "grad_norm": 0.572250485420227, + "learning_rate": 1.6536966062249932e-06, + "loss": 0.324951171875, + "step": 1830, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.889588998430743 + }, + { + "epoch": 0.7437430337420204, + "grad_norm": 0.42056840658187866, + "learning_rate": 1.6293822215888533e-06, + "loss": 0.3470703125, + "step": 1835, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8832219765723289 + }, + { + "epoch": 0.7457695815178843, + "grad_norm": 0.45901891589164734, + "learning_rate": 1.6052130689481948e-06, + "loss": 0.365869140625, + "step": 1840, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8776322582398174 + }, + { + "epoch": 0.7477961292937481, + "grad_norm": 0.4727613627910614, + "learning_rate": 1.5811901896949788e-06, + "loss": 0.3380126953125, + "step": 1845, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8879822377210093 + }, + { + "epoch": 0.7498226770696119, + "grad_norm": 0.24706824123859406, + "learning_rate": 1.5573146189185945e-06, + "loss": 0.3182373046875, + "step": 1850, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8955709839597797 + }, + { + "epoch": 0.7518492248454758, + "grad_norm": 0.41805166006088257, + "learning_rate": 1.533587385361252e-06, + "loss": 0.3082763671875, + "step": 1855, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8968251179894365 + }, + { + "epoch": 0.7538757726213395, + "grad_norm": 0.38584384322166443, + "learning_rate": 1.5100095113736635e-06, + "loss": 0.324169921875, + "step": 1860, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8908545262919716 + }, + { + "epoch": 0.7559023203972034, + "grad_norm": 0.4495343863964081, + "learning_rate": 1.4865820128709812e-06, + "loss": 0.34052734375, + "step": 1865, + "tIoU": 0.99375, + "tRecall_30": 0.99375, + "tRecall_50": 0.99375, + "tRecall_70": 0.99375, + "token_acc": 0.8863164086726897 + }, + { + "epoch": 0.7579288681730671, + "grad_norm": 0.40899696946144104, + "learning_rate": 1.4633058992890387e-06, + "loss": 0.3251708984375, + "step": 1870, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8894297164829809 + }, + { + "epoch": 0.759955415948931, + "grad_norm": 0.30098530650138855, + "learning_rate": 1.4401821735408456e-06, + "loss": 0.34287109375, + "step": 1875, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8844233449875459 + }, + { + "epoch": 0.7619819637247948, + "grad_norm": 0.392616868019104, + "learning_rate": 1.4172118319733796e-06, + "loss": 0.297802734375, + "step": 1880, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.9007301045196695 + }, + { + "epoch": 0.7640085115006586, + "grad_norm": 0.46458375453948975, + "learning_rate": 1.3943958643246514e-06, + "loss": 0.341845703125, + "step": 1885, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8854414186853464 + }, + { + "epoch": 0.7660350592765225, + "grad_norm": 0.4317547082901001, + "learning_rate": 1.3717352536810679e-06, + "loss": 0.3319580078125, + "step": 1890, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8870785499809603 + }, + { + "epoch": 0.7680616070523862, + "grad_norm": 0.48436179757118225, + "learning_rate": 1.3492309764350597e-06, + "loss": 0.337744140625, + "step": 1895, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8858699511332313 + }, + { + "epoch": 0.7700881548282501, + "grad_norm": 0.41826820373535156, + "learning_rate": 1.3268840022430306e-06, + "loss": 0.3273193359375, + "step": 1900, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.890079424999245 + }, + { + "epoch": 0.7721147026041139, + "grad_norm": 0.5227497220039368, + "learning_rate": 1.3046952939835545e-06, + "loss": 0.350390625, + "step": 1905, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8822442209117562 + }, + { + "epoch": 0.7741412503799777, + "grad_norm": 0.412296861410141, + "learning_rate": 1.2826658077159054e-06, + "loss": 0.328564453125, + "step": 1910, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.888492897475502 + }, + { + "epoch": 0.7761677981558415, + "grad_norm": 0.47801196575164795, + "learning_rate": 1.2607964926388494e-06, + "loss": 0.32509765625, + "step": 1915, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8916572717023675 + }, + { + "epoch": 0.7781943459317053, + "grad_norm": 0.41708609461784363, + "learning_rate": 1.2390882910497627e-06, + "loss": 0.325634765625, + "step": 1920, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8909982773219656 + }, + { + "epoch": 0.7802208937075692, + "grad_norm": 0.39932745695114136, + "learning_rate": 1.2175421383040088e-06, + "loss": 0.3413818359375, + "step": 1925, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8852147296169186 + }, + { + "epoch": 0.782247441483433, + "grad_norm": 0.3740224838256836, + "learning_rate": 1.1961589627746566e-06, + "loss": 0.3327392578125, + "step": 1930, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.888258064516129 + }, + { + "epoch": 0.7842739892592968, + "grad_norm": 0.4308731257915497, + "learning_rate": 1.1749396858124611e-06, + "loss": 0.3427490234375, + "step": 1935, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8836409967702746 + }, + { + "epoch": 0.7863005370351606, + "grad_norm": 0.43539366126060486, + "learning_rate": 1.1538852217061853e-06, + "loss": 0.34677734375, + "step": 1940, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8837174094771694 + }, + { + "epoch": 0.7883270848110244, + "grad_norm": 0.40453237295150757, + "learning_rate": 1.1329964776431812e-06, + "loss": 0.312109375, + "step": 1945, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8960424425305664 + }, + { + "epoch": 0.7903536325868883, + "grad_norm": 0.46798184514045715, + "learning_rate": 1.112274353670323e-06, + "loss": 0.3400146484375, + "step": 1950, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8842769195905881 + }, + { + "epoch": 0.792380180362752, + "grad_norm": 0.5182797908782959, + "learning_rate": 1.0917197426552089e-06, + "loss": 0.33603515625, + "step": 1955, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8882436625418512 + }, + { + "epoch": 0.7944067281386159, + "grad_norm": 0.40419793128967285, + "learning_rate": 1.0713335302477073e-06, + "loss": 0.3489990234375, + "step": 1960, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.882445557326049 + }, + { + "epoch": 0.7964332759144797, + "grad_norm": 0.47690391540527344, + "learning_rate": 1.0511165948417774e-06, + "loss": 0.351708984375, + "step": 1965, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8830510993235429 + }, + { + "epoch": 0.7984598236903435, + "grad_norm": 0.35004881024360657, + "learning_rate": 1.0310698075376368e-06, + "loss": 0.30078125, + "step": 1970, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.9015132509236875 + }, + { + "epoch": 0.8004863714662073, + "grad_norm": 0.43832194805145264, + "learning_rate": 1.0111940321042146e-06, + "loss": 0.3324951171875, + "step": 1975, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8875856153256306 + }, + { + "epoch": 0.8025129192420711, + "grad_norm": 0.33005771040916443, + "learning_rate": 9.914901249419512e-07, + "loss": 0.3274658203125, + "step": 1980, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8900861560250206 + }, + { + "epoch": 0.804539467017935, + "grad_norm": 0.2617853581905365, + "learning_rate": 9.719589350458753e-07, + "loss": 0.3199951171875, + "step": 1985, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8958106699301968 + }, + { + "epoch": 0.8065660147937987, + "grad_norm": 0.5502761602401733, + "learning_rate": 9.526013039690429e-07, + "loss": 0.3506591796875, + "step": 1990, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.880904348734421 + }, + { + "epoch": 0.8085925625696626, + "grad_norm": 0.4234298765659332, + "learning_rate": 9.334180657862629e-07, + "loss": 0.3437744140625, + "step": 1995, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8829914034772302 + }, + { + "epoch": 0.8106191103455264, + "grad_norm": 0.4022880494594574, + "learning_rate": 9.144100470581718e-07, + "loss": 0.3411865234375, + "step": 2000, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.884309301945398 + }, + { + "epoch": 0.8126456581213902, + "grad_norm": 0.41832342743873596, + "learning_rate": 8.955780667956032e-07, + "loss": 0.338232421875, + "step": 2005, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8855404089581305 + }, + { + "epoch": 0.814672205897254, + "grad_norm": 0.4092077612876892, + "learning_rate": 8.769229364243137e-07, + "loss": 0.3219970703125, + "step": 2010, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.891176808231676 + }, + { + "epoch": 0.8166987536731178, + "grad_norm": 0.4354773759841919, + "learning_rate": 8.584454597500092e-07, + "loss": 0.3410888671875, + "step": 2015, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8857620386816657 + }, + { + "epoch": 0.8187253014489817, + "grad_norm": 0.49091804027557373, + "learning_rate": 8.401464329237185e-07, + "loss": 0.3350830078125, + "step": 2020, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8864969363347422 + }, + { + "epoch": 0.8207518492248455, + "grad_norm": 0.4098585546016693, + "learning_rate": 8.220266444074853e-07, + "loss": 0.3209716796875, + "step": 2025, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8926854374810778 + }, + { + "epoch": 0.8227783970007093, + "grad_norm": 0.3522760272026062, + "learning_rate": 8.040868749403968e-07, + "loss": 0.314892578125, + "step": 2030, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8934764299338601 + }, + { + "epoch": 0.8248049447765731, + "grad_norm": 0.40749940276145935, + "learning_rate": 7.863278975049388e-07, + "loss": 0.3139892578125, + "step": 2035, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8930227973747072 + }, + { + "epoch": 0.8268314925524369, + "grad_norm": 0.4043844938278198, + "learning_rate": 7.687504772936983e-07, + "loss": 0.3261474609375, + "step": 2040, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8905281649048885 + }, + { + "epoch": 0.8288580403283008, + "grad_norm": 0.4920687973499298, + "learning_rate": 7.513553716763849e-07, + "loss": 0.3332763671875, + "step": 2045, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8874587937196271 + }, + { + "epoch": 0.8308845881041645, + "grad_norm": 0.32292449474334717, + "learning_rate": 7.341433301672047e-07, + "loss": 0.2773193359375, + "step": 2050, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.9084233121903194 + }, + { + "epoch": 0.8329111358800284, + "grad_norm": 0.3950773775577545, + "learning_rate": 7.171150943925565e-07, + "loss": 0.3315673828125, + "step": 2055, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8875533463255791 + }, + { + "epoch": 0.8349376836558922, + "grad_norm": 0.43536999821662903, + "learning_rate": 7.002713980590864e-07, + "loss": 0.340283203125, + "step": 2060, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8850941570345794 + }, + { + "epoch": 0.836964231431756, + "grad_norm": 0.47620904445648193, + "learning_rate": 6.836129669220687e-07, + "loss": 0.3458251953125, + "step": 2065, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8832007379222876 + }, + { + "epoch": 0.8389907792076198, + "grad_norm": 0.32710176706314087, + "learning_rate": 6.671405187541369e-07, + "loss": 0.32021484375, + "step": 2070, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8921389175833802 + }, + { + "epoch": 0.8410173269834836, + "grad_norm": 0.3626502752304077, + "learning_rate": 6.508547633143519e-07, + "loss": 0.319482421875, + "step": 2075, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8915977638033111 + }, + { + "epoch": 0.8430438747593475, + "grad_norm": 0.4492169916629791, + "learning_rate": 6.347564023176283e-07, + "loss": 0.3444580078125, + "step": 2080, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8843687264694935 + }, + { + "epoch": 0.8450704225352113, + "grad_norm": 0.4278513193130493, + "learning_rate": 6.188461294044923e-07, + "loss": 0.3422119140625, + "step": 2085, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8840360895664138 + }, + { + "epoch": 0.8470969703110751, + "grad_norm": 0.4415869414806366, + "learning_rate": 6.031246301111987e-07, + "loss": 0.347705078125, + "step": 2090, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8835707624529697 + }, + { + "epoch": 0.8491235180869389, + "grad_norm": 0.3770047724246979, + "learning_rate": 5.875925818401873e-07, + "loss": 0.3265625, + "step": 2095, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8900356756694248 + }, + { + "epoch": 0.8511500658628027, + "grad_norm": 0.37200838327407837, + "learning_rate": 5.722506538309025e-07, + "loss": 0.3359619140625, + "step": 2100, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8867193743928979 + }, + { + "epoch": 0.8531766136386665, + "grad_norm": 0.4725235402584076, + "learning_rate": 5.570995071309515e-07, + "loss": 0.346875, + "step": 2105, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8840582633957392 + }, + { + "epoch": 0.8552031614145303, + "grad_norm": 0.3805801272392273, + "learning_rate": 5.421397945676255e-07, + "loss": 0.32822265625, + "step": 2110, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8886891705894553 + }, + { + "epoch": 0.8572297091903942, + "grad_norm": 0.44028645753860474, + "learning_rate": 5.273721607197651e-07, + "loss": 0.3368408203125, + "step": 2115, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8858809924306139 + }, + { + "epoch": 0.859256256966258, + "grad_norm": 0.3674177825450897, + "learning_rate": 5.127972418899941e-07, + "loss": 0.3289794921875, + "step": 2120, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8882873734649193 + }, + { + "epoch": 0.8612828047421218, + "grad_norm": 0.4421343505382538, + "learning_rate": 4.984156660772976e-07, + "loss": 0.3440185546875, + "step": 2125, + "tIoU": 0.9989355087280274, + "tRecall_30": 0.9997768402099609, + "tRecall_50": 0.9997768402099609, + "tRecall_70": 0.9997768402099609, + "token_acc": 0.883937518420277 + }, + { + "epoch": 0.8633093525179856, + "grad_norm": 0.43405815958976746, + "learning_rate": 4.842280529499671e-07, + "loss": 0.351904296875, + "step": 2130, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8818160076530612 + }, + { + "epoch": 0.8653359002938494, + "grad_norm": 0.37876492738723755, + "learning_rate": 4.7023501381889326e-07, + "loss": 0.30791015625, + "step": 2135, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8961273254511662 + }, + { + "epoch": 0.8673624480697133, + "grad_norm": 0.425828218460083, + "learning_rate": 4.5643715161123424e-07, + "loss": 0.3364501953125, + "step": 2140, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8863891502546964 + }, + { + "epoch": 0.869388995845577, + "grad_norm": 0.4508116841316223, + "learning_rate": 4.428350608444337e-07, + "loss": 0.34580078125, + "step": 2145, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8834513383656141 + }, + { + "epoch": 0.8714155436214409, + "grad_norm": 0.3458070158958435, + "learning_rate": 4.2942932760060395e-07, + "loss": 0.3091552734375, + "step": 2150, + "tIoU": 0.9999964714050293, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8950527958034684 + }, + { + "epoch": 0.8734420913973047, + "grad_norm": 0.5336869359016418, + "learning_rate": 4.1622052950126946e-07, + "loss": 0.35849609375, + "step": 2155, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8784318039617439 + }, + { + "epoch": 0.8754686391731685, + "grad_norm": 0.35457611083984375, + "learning_rate": 4.0320923568248796e-07, + "loss": 0.3381591796875, + "step": 2160, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8862457701347886 + }, + { + "epoch": 0.8774951869490323, + "grad_norm": 0.3743188679218292, + "learning_rate": 3.903960067703177e-07, + "loss": 0.3113525390625, + "step": 2165, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8935743921322947 + }, + { + "epoch": 0.8795217347248961, + "grad_norm": 0.4452930688858032, + "learning_rate": 3.7778139485667055e-07, + "loss": 0.353515625, + "step": 2170, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8812360435248409 + }, + { + "epoch": 0.88154828250076, + "grad_norm": 0.4194066524505615, + "learning_rate": 3.653659434755125e-07, + "loss": 0.3264892578125, + "step": 2175, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8891549098383631 + }, + { + "epoch": 0.8835748302766238, + "grad_norm": 0.4035830497741699, + "learning_rate": 3.531501875794552e-07, + "loss": 0.3505615234375, + "step": 2180, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8810427993168125 + }, + { + "epoch": 0.8856013780524876, + "grad_norm": 0.37540069222450256, + "learning_rate": 3.411346535166998e-07, + "loss": 0.345849609375, + "step": 2185, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8822157717672492 + }, + { + "epoch": 0.8876279258283514, + "grad_norm": 0.4143151044845581, + "learning_rate": 3.293198590083613e-07, + "loss": 0.33671875, + "step": 2190, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8863002774610664 + }, + { + "epoch": 0.8896544736042152, + "grad_norm": 0.44223761558532715, + "learning_rate": 3.1770631312615483e-07, + "loss": 0.3336669921875, + "step": 2195, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8871216076018793 + }, + { + "epoch": 0.891681021380079, + "grad_norm": 0.3474159836769104, + "learning_rate": 3.0629451627047e-07, + "loss": 0.312890625, + "step": 2200, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8945662581101734 + }, + { + "epoch": 0.8937075691559428, + "grad_norm": 0.3162406086921692, + "learning_rate": 2.9508496014880205e-07, + "loss": 0.328857421875, + "step": 2205, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8887455197132617 + }, + { + "epoch": 0.8957341169318067, + "grad_norm": 0.4361012279987335, + "learning_rate": 2.840781277545712e-07, + "loss": 0.3455322265625, + "step": 2210, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.885535628989535 + }, + { + "epoch": 0.8977606647076705, + "grad_norm": 0.41882529854774475, + "learning_rate": 2.732744933463055e-07, + "loss": 0.3302978515625, + "step": 2215, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8884911839624202 + }, + { + "epoch": 0.8997872124835343, + "grad_norm": 0.48173168301582336, + "learning_rate": 2.626745224272126e-07, + "loss": 0.3207763671875, + "step": 2220, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8912699949407569 + }, + { + "epoch": 0.9018137602593981, + "grad_norm": 0.3567911982536316, + "learning_rate": 2.5227867172511746e-07, + "loss": 0.33603515625, + "step": 2225, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8863890027430085 + }, + { + "epoch": 0.9038403080352619, + "grad_norm": 0.4256726801395416, + "learning_rate": 2.4208738917278654e-07, + "loss": 0.303759765625, + "step": 2230, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8972713804085347 + }, + { + "epoch": 0.9058668558111257, + "grad_norm": 0.49550139904022217, + "learning_rate": 2.3210111388862312e-07, + "loss": 0.34375, + "step": 2235, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8858439106262457 + }, + { + "epoch": 0.9078934035869896, + "grad_norm": 0.38924527168273926, + "learning_rate": 2.2232027615775088e-07, + "loss": 0.3441650390625, + "step": 2240, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.884348966874385 + }, + { + "epoch": 0.9099199513628534, + "grad_norm": 0.38762596249580383, + "learning_rate": 2.127452974134725e-07, + "loss": 0.32607421875, + "step": 2245, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8882261345656378 + }, + { + "epoch": 0.9119464991387172, + "grad_norm": 0.520622193813324, + "learning_rate": 2.0337659021911094e-07, + "loss": 0.3317138671875, + "step": 2250, + "tIoU": 0.9982122421264649, + "tRecall_30": 0.999609375, + "tRecall_50": 0.999609375, + "tRecall_70": 0.999609375, + "token_acc": 0.8874109294771036 + }, + { + "epoch": 0.913973046914581, + "grad_norm": 0.5936887264251709, + "learning_rate": 1.9421455825023194e-07, + "loss": 0.3436767578125, + "step": 2255, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8838248321372174 + }, + { + "epoch": 0.9159995946904448, + "grad_norm": 0.526875376701355, + "learning_rate": 1.8525959627725188e-07, + "loss": 0.3408447265625, + "step": 2260, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8860013722851187 + }, + { + "epoch": 0.9180261424663086, + "grad_norm": 0.444475382566452, + "learning_rate": 1.7651209014842963e-07, + "loss": 0.3270751953125, + "step": 2265, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8931827257687261 + }, + { + "epoch": 0.9200526902421725, + "grad_norm": 0.34051868319511414, + "learning_rate": 1.6797241677323773e-07, + "loss": 0.3276611328125, + "step": 2270, + "tIoU": 0.9999456405639648, + "tRecall_30": 0.9999431610107422, + "tRecall_50": 0.9999431610107422, + "tRecall_70": 0.9999431610107422, + "token_acc": 0.8906207714215756 + }, + { + "epoch": 0.9220792380180363, + "grad_norm": 0.40930411219596863, + "learning_rate": 1.5964094410612418e-07, + "loss": 0.3405029296875, + "step": 2275, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8846896342196707 + }, + { + "epoch": 0.9241057857939001, + "grad_norm": 0.35660627484321594, + "learning_rate": 1.5151803113065854e-07, + "loss": 0.3163818359375, + "step": 2280, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.893365429779239 + }, + { + "epoch": 0.9261323335697639, + "grad_norm": 0.48520004749298096, + "learning_rate": 1.4360402784406202e-07, + "loss": 0.3478759765625, + "step": 2285, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8817491056381752 + }, + { + "epoch": 0.9281588813456277, + "grad_norm": 0.4974427819252014, + "learning_rate": 1.3589927524213232e-07, + "loss": 0.3416748046875, + "step": 2290, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8828711924600464 + }, + { + "epoch": 0.9301854291214915, + "grad_norm": 0.4171598553657532, + "learning_rate": 1.2840410530454262e-07, + "loss": 0.3143310546875, + "step": 2295, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8934846058623783 + }, + { + "epoch": 0.9322119768973554, + "grad_norm": 0.3226058781147003, + "learning_rate": 1.211188409805447e-07, + "loss": 0.3171875, + "step": 2300, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8943094746978785 + }, + { + "epoch": 0.9342385246732192, + "grad_norm": 0.443516343832016, + "learning_rate": 1.1404379617504846e-07, + "loss": 0.318310546875, + "step": 2305, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.892321063327866 + }, + { + "epoch": 0.936265072449083, + "grad_norm": 0.45417943596839905, + "learning_rate": 1.0717927573510312e-07, + "loss": 0.331787109375, + "step": 2310, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8895140322354761 + }, + { + "epoch": 0.9382916202249468, + "grad_norm": 0.4532037675380707, + "learning_rate": 1.0052557543675346e-07, + "loss": 0.33115234375, + "step": 2315, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8877487976546694 + }, + { + "epoch": 0.9403181680008106, + "grad_norm": 0.43640556931495667, + "learning_rate": 9.40829819723027e-08, + "loss": 0.318115234375, + "step": 2320, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.894122094364284 + }, + { + "epoch": 0.9423447157766744, + "grad_norm": 0.39802286028862, + "learning_rate": 8.785177293795577e-08, + "loss": 0.3271728515625, + "step": 2325, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8903346784666747 + }, + { + "epoch": 0.9443712635525382, + "grad_norm": 0.35078153014183044, + "learning_rate": 8.183221682186048e-08, + "loss": 0.31474609375, + "step": 2330, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8940282315868322 + }, + { + "epoch": 0.9463978113284021, + "grad_norm": 0.4221342206001282, + "learning_rate": 7.602457299253685e-08, + "loss": 0.33984375, + "step": 2335, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8851874961003307 + }, + { + "epoch": 0.9484243591042659, + "grad_norm": 0.46918657422065735, + "learning_rate": 7.042909168770373e-08, + "loss": 0.3451416015625, + "step": 2340, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.884408887239176 + }, + { + "epoch": 0.9504509068801297, + "grad_norm": 0.2663903534412384, + "learning_rate": 6.504601400349419e-08, + "loss": 0.3374267578125, + "step": 2345, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8878492824179726 + }, + { + "epoch": 0.9524774546559935, + "grad_norm": 0.43997228145599365, + "learning_rate": 5.987557188407156e-08, + "loss": 0.319580078125, + "step": 2350, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8926250406178918 + }, + { + "epoch": 0.9545040024318573, + "grad_norm": 0.37927335500717163, + "learning_rate": 5.491798811162852e-08, + "loss": 0.33017578125, + "step": 2355, + "tIoU": 0.9998649597167969, + "tRecall_30": 0.9998437881469726, + "tRecall_50": 0.9998437881469726, + "tRecall_70": 0.9998437881469726, + "token_acc": 0.8897617369237276 + }, + { + "epoch": 0.9565305502077212, + "grad_norm": 0.4116072654724121, + "learning_rate": 5.0173476296795363e-08, + "loss": 0.3291748046875, + "step": 2360, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8886311682565646 + }, + { + "epoch": 0.958557097983585, + "grad_norm": 0.4057512581348419, + "learning_rate": 4.564224086943014e-08, + "loss": 0.343798828125, + "step": 2365, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8851102244892806 + }, + { + "epoch": 0.9605836457594488, + "grad_norm": 0.4532250463962555, + "learning_rate": 4.132447706981624e-08, + "loss": 0.3299072265625, + "step": 2370, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.889131115705068 + }, + { + "epoch": 0.9626101935353126, + "grad_norm": 0.3738594651222229, + "learning_rate": 3.722037094024244e-08, + "loss": 0.315185546875, + "step": 2375, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8929597656371739 + }, + { + "epoch": 0.9646367413111764, + "grad_norm": 0.3902566432952881, + "learning_rate": 3.333009931699438e-08, + "loss": 0.330908203125, + "step": 2380, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8876850592640813 + }, + { + "epoch": 0.9666632890870402, + "grad_norm": 0.4984258711338043, + "learning_rate": 2.9653829822728353e-08, + "loss": 0.3281982421875, + "step": 2385, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8927605491760068 + }, + { + "epoch": 0.968689836862904, + "grad_norm": 0.3534846305847168, + "learning_rate": 2.6191720859254922e-08, + "loss": 0.317041015625, + "step": 2390, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8924771605711797 + }, + { + "epoch": 0.9707163846387679, + "grad_norm": 0.4049256145954132, + "learning_rate": 2.2943921600708796e-08, + "loss": 0.3455322265625, + "step": 2395, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8843115851136679 + }, + { + "epoch": 0.9727429324146317, + "grad_norm": 0.3752588927745819, + "learning_rate": 1.9910571987125094e-08, + "loss": 0.3220458984375, + "step": 2400, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8901173485569299 + }, + { + "epoch": 0.9747694801904955, + "grad_norm": 0.4580046534538269, + "learning_rate": 1.709180271840638e-08, + "loss": 0.329150390625, + "step": 2405, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8885664172418579 + }, + { + "epoch": 0.9767960279663593, + "grad_norm": 0.35217657685279846, + "learning_rate": 1.4487735248692737e-08, + "loss": 0.3216552734375, + "step": 2410, + "tIoU": 0.9999425888061524, + "tRecall_30": 0.9999431610107422, + "tRecall_50": 0.9999431610107422, + "tRecall_70": 0.9999431610107422, + "token_acc": 0.8933333333333333 + }, + { + "epoch": 0.9788225757422231, + "grad_norm": 0.41408440470695496, + "learning_rate": 1.2098481781130377e-08, + "loss": 0.3031494140625, + "step": 2415, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8987533738838125 + }, + { + "epoch": 0.980849123518087, + "grad_norm": 0.34325486421585083, + "learning_rate": 9.924145263033313e-09, + "loss": 0.3330322265625, + "step": 2420, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.887853635954646 + }, + { + "epoch": 0.9828756712939507, + "grad_norm": 0.379108190536499, + "learning_rate": 7.964819381449662e-09, + "loss": 0.310400390625, + "step": 2425, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8947683743455007 + }, + { + "epoch": 0.9849022190698146, + "grad_norm": 0.46881887316703796, + "learning_rate": 6.22058855912544e-09, + "loss": 0.3291748046875, + "step": 2430, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.889718241829013 + }, + { + "epoch": 0.9869287668456784, + "grad_norm": 0.38360169529914856, + "learning_rate": 4.691527950865804e-09, + "loss": 0.3313232421875, + "step": 2435, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8881468006851406 + }, + { + "epoch": 0.9889553146215422, + "grad_norm": 0.5358332395553589, + "learning_rate": 3.3777034402981966e-09, + "loss": 0.3405029296875, + "step": 2440, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8851056035571725 + }, + { + "epoch": 0.990981862397406, + "grad_norm": 0.3311910033226013, + "learning_rate": 2.2791716370307305e-09, + "loss": 0.3314453125, + "step": 2445, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8888736476734616 + }, + { + "epoch": 0.9930084101732698, + "grad_norm": 0.4242856800556183, + "learning_rate": 1.3959798742169128e-09, + "loss": 0.3162109375, + "step": 2450, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8925040315518001 + }, + { + "epoch": 0.9950349579491337, + "grad_norm": 0.49177971482276917, + "learning_rate": 7.28166206512837e-10, + "loss": 0.334423828125, + "step": 2455, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8867867609032926 + }, + { + "epoch": 0.9970615057249975, + "grad_norm": 0.5094365477561951, + "learning_rate": 2.7575940843960204e-10, + "loss": 0.3192626953125, + "step": 2460, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8922791595741978 + }, + { + "epoch": 0.9990880535008613, + "grad_norm": 0.3633192181587219, + "learning_rate": 3.877897314263912e-11, + "loss": 0.3603271484375, + "step": 2465, + "tIoU": 1.0, + "tRecall_30": 1.0, + "tRecall_50": 1.0, + "tRecall_70": 1.0, + "token_acc": 0.8784863668900892 + } + ], + "logging_steps": 5, + "max_steps": 2468, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 8.412078391672858e+19, + "train_batch_size": 4, + "trial_name": null, + "trial_params": null +} diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/training_args.bin b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..8861411424bd302439bea49d30babdbc7a2a75f5 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:50a3deaef62d81116f9e8987aa53c636b68dd24dc259869b747f0f44a3e9a257 +size 9297 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/video_preprocessor_config.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/video_preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..e32b1d90356f74457edbfca27aa1e121dab7cdc7 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/video_preprocessor_config.json @@ -0,0 +1,41 @@ +{ + "crop_size": null, + "data_format": "channels_first", + "default_to_square": true, + "device": null, + "do_center_crop": null, + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "input_data_format": null, + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "num_frames": null, + "pad_size": null, + "patch_size": 16, + "processor_class": "Qwen3VLProcessor", + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_metadata": null, + "video_processor_type": "Qwen3VLVideoProcessor" +} diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/zero_to_fp32.py b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..5995d6e6f04e43b989587aa9022a3aef0c66d694 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468/zero_to_fp32.py @@ -0,0 +1,760 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: +# python zero_to_fp32.py . output_dir/ +# or +# python zero_to_fp32.py . output_dir/ --safe_serialization + +import argparse +import torch +import glob +import math +import os +import re +import gc +import json +import numpy as np +from tqdm import tqdm +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device, weights_only=False) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + total_files = len(files) + state_dicts = [] + for f in tqdm(files, desc='Loading checkpoint shards'): + state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if ZERO_STAGE not in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +class GatheredTensor: + """ + A pseudo tensor that collects partitioned weights. + It is more memory efficient when there are multiple groups. + """ + + def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape): + self.flat_groups = flat_groups + self.flat_groups_offset = flat_groups_offset + self.offset = offset + self.partitioned_numel = partitioned_numel + self.shape = shape + self.dtype = self.flat_groups[0][0].dtype + + def contiguous(self): + """ + Merge partitioned weights from flat_groups into a single tensor. + """ + end_idx = self.offset + self.partitioned_numel + world_size = len(self.flat_groups) + pad_flat_param_chunks = [] + + for rank_i in range(world_size): + # for each rank, we need to collect weights from related group/groups + flat_groups_at_rank_i = self.flat_groups[rank_i] + start_group_id = None + end_group_id = None + for group_id in range(len(self.flat_groups_offset)): + if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]: + start_group_id = group_id + if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]: + end_group_id = group_id + break + # collect weights from related group/groups + for group_id in range(start_group_id, end_group_id + 1): + flat_tensor = flat_groups_at_rank_i[group_id] + start_offset = self.offset - self.flat_groups_offset[group_id] + end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id] + pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset]) + + # collect weights from all ranks + pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0) + param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous() + return param + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size + + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]])) + for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'): + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # memory efficient tensor + tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape) + state_dict[name] = tensor + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def to_torch_tensor(state_dict, return_empty_tensor=False): + """ + Convert state_dict of GatheredTensor to torch tensor + """ + torch_state_dict = {} + converted_tensors = {} + for name, tensor in state_dict.items(): + tensor_id = id(tensor) + if tensor_id in converted_tensors: # shared tensors + shared_tensor = torch_state_dict[converted_tensors[tensor_id]] + torch_state_dict[name] = shared_tensor + else: + converted_tensors[tensor_id] = name + if return_empty_tensor: + torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype) + else: + torch_state_dict[name] = tensor.contiguous() + return torch_state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag=None, + exclude_frozen_parameters=False, + lazy_mode=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient. + Convert the pesduo tensor to torch tensor by ``.contiguous()`` + + Returns: + - pytorch ``state_dict`` + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + Note: the above usage may not work if your application doesn't have sufficient free CPU memory. + You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. Or you can load state_dict in lazy mode :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu + for name, lazy_tensor in state_dict.item(): + tensor = lazy_tensor.contiguous() # to cpu + print(name, tensor) + # del tensor to release memory if it no longer in use + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + if lazy_mode: + return state_dict + else: + return to_torch_tensor(state_dict) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, + output_dir, + max_shard_size="5GB", + safe_serialization=False, + tag=None, + exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_dir``: directory to the pytorch fp32 state_dict output files + - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB + - ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`). + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + # Dependency pre-check + if safe_serialization: + try: + from safetensors.torch import save_file + except ImportError: + print('If you want to use `safe_serialization`, please `pip install safetensors`') + raise + if max_shard_size is not None: + try: + from huggingface_hub import split_torch_state_dict_into_shards + except ImportError: + print('If you want to use `max_shard_size`, please `pip install huggingface_hub`') + raise + + # Convert zero checkpoint to state_dict + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag, + exclude_frozen_parameters, + lazy_mode=True) + + # Shard the model if it is too big. + weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin" + if max_shard_size is not None: + filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors") + # an memory-efficient approach for sharding + empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True) + state_dict_split = split_torch_state_dict_into_shards(empty_state_dict, + filename_pattern=filename_pattern, + max_shard_size=max_shard_size) + else: + from collections import namedtuple + StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"]) + state_dict_split = StateDictSplit(is_sharded=False, + filename_to_tensors={weights_name: list(state_dict.keys())}) + + # Save the model by shard + os.makedirs(output_dir, exist_ok=True) + filename_to_tensors = state_dict_split.filename_to_tensors.items() + for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"): + shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors} + shard_state_dict = to_torch_tensor(shard_state_dict) + output_path = os.path.join(output_dir, shard_file) + if safe_serialization: + save_file(shard_state_dict, output_path, metadata={"format": "pt"}) + else: + torch.save(shard_state_dict, output_path) + # release the memory of current shard + for tensor_name in list(shard_state_dict.keys()): + del state_dict[tensor_name] + del shard_state_dict[tensor_name] + del shard_state_dict + gc.collect() + + # Save index if sharded + if state_dict_split.is_sharded: + index = { + "metadata": state_dict_split.metadata, + "weight_map": state_dict_split.tensor_to_filename, + } + save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json" + save_index_file = os.path.join(output_dir, save_index_file) + with open(save_index_file, "w", encoding="utf-8") as f: + content = json.dumps(index, indent=2, sort_keys=True) + "\n" + f.write(content) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info("Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info("Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument("output_dir", + type=str, + help="directory to the pytorch fp32 state_dict output files" + "(e.g. path/checkpoint-12-output/)") + parser.add_argument( + "--max_shard_size", + type=str, + default="5GB", + help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size" + "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`" + "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances" + "without CPU OOM issues.") + parser.add_argument( + "--safe_serialization", + default=False, + action='store_true', + help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_dir, + max_shard_size=args.max_shard_size, + safe_serialization=args.safe_serialization, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/logging.jsonl b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/logging.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..df94253c2019051d924bc54d3ae0a2c1aa1f565e --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/logging.jsonl @@ -0,0 +1,496 @@ +{"loss": 0.703125, "grad_norm": 1.20598018, "learning_rate": 1.3e-07, "tIoU": 0.25165161, "tRecall_30": 0.78593755, "tRecall_50": 0.56979173, "tRecall_70": 0.54635423, "token_acc": 0.81051313, "epoch": 0.00040531, "global_step/max_steps": "1/2468", "percentage": "0.04%", "elapsed_time": "3m 29s", "remaining_time": "5d 23h 44m 42s", "memory(GiB)": 37.01, "train_speed(iter/s)": 0.004767} +{"loss": 0.73364258, "grad_norm": 1.4883641, "learning_rate": 6.7e-07, "tIoU": 0.23187013, "tRecall_30": 0.74438709, "tRecall_50": 0.61836988, "tRecall_70": 0.52761477, "token_acc": 0.80650312, "epoch": 0.00202655, "global_step/max_steps": "5/2468", "percentage": "0.20%", "elapsed_time": "5m 37s", "remaining_time": "1d 22h 9m 7s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.014824} +{"loss": 0.74443359, "grad_norm": 1.14006388, "learning_rate": 1.33e-06, "tIoU": 0.25241461, "tRecall_30": 0.78672667, "tRecall_50": 0.64610157, "tRecall_70": 0.51099739, "token_acc": 0.80266414, "epoch": 0.0040531, "global_step/max_steps": "10/2468", "percentage": "0.41%", "elapsed_time": "8m 12s", "remaining_time": "1d 9h 36m 24s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.020317} +{"loss": 0.67539063, "grad_norm": 0.88751131, "learning_rate": 2e-06, "tIoU": 0.28452642, "tRecall_30": 0.84657736, "tRecall_50": 0.7617857, "tRecall_70": 0.72272325, "token_acc": 0.81187419, "epoch": 0.00607964, "global_step/max_steps": "15/2468", "percentage": "0.61%", "elapsed_time": "10m 9s", "remaining_time": "1d 3h 41m 29s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.024606} +{"loss": 0.57397461, "grad_norm": 1.01068974, "learning_rate": 2.67e-06, "tIoU": 0.24748538, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.82947788, "epoch": 0.00810619, "global_step/max_steps": "20/2468", "percentage": "0.81%", "elapsed_time": "11m 16s", "remaining_time": "23h 0m 19s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.029558} +{"loss": 0.47485352, "grad_norm": 0.73120511, "learning_rate": 3.33e-06, "tIoU": 0.84586229, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.85256984, "epoch": 0.01013274, "global_step/max_steps": "25/2468", "percentage": "1.01%", "elapsed_time": "12m 26s", "remaining_time": "20h 15m 3s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.03351} +{"loss": 0.42241211, "grad_norm": 0.48748231, "learning_rate": 4e-06, "tIoU": 0.99243021, "tRecall_30": 0.98937492, "tRecall_50": 0.98937492, "tRecall_70": 0.98937492, "token_acc": 0.86281511, "epoch": 0.01215929, "global_step/max_steps": "30/2468", "percentage": "1.22%", "elapsed_time": "13m 33s", "remaining_time": "18h 22m 1s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.036872} +{"loss": 0.40251465, "grad_norm": 0.55686867, "learning_rate": 4.67e-06, "tIoU": 0.99562025, "tRecall_30": 0.9938839, "tRecall_50": 0.9938839, "tRecall_70": 0.9938839, "token_acc": 0.86901144, "epoch": 0.01418583, "global_step/max_steps": "35/2468", "percentage": "1.42%", "elapsed_time": "14m 41s", "remaining_time": "17h 0m 48s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.039723} +{"loss": 0.43410645, "grad_norm": 0.62369204, "learning_rate": 5.33e-06, "tIoU": 0.99956207, "tRecall_30": 0.99874992, "tRecall_50": 0.99874992, "tRecall_70": 0.99874992, "token_acc": 0.85771572, "epoch": 0.01621238, "global_step/max_steps": "40/2468", "percentage": "1.62%", "elapsed_time": "15m 47s", "remaining_time": "15h 58m 35s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.042215} +{"loss": 0.3864502, "grad_norm": 0.60356069, "learning_rate": 6e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87276995, "epoch": 0.01823893, "global_step/max_steps": "45/2468", "percentage": "1.82%", "elapsed_time": "16m 56s", "remaining_time": "15h 12m 24s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.04426} +{"loss": 0.3807373, "grad_norm": 0.44803467, "learning_rate": 6.67e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87473249, "epoch": 0.02026548, "global_step/max_steps": "50/2468", "percentage": "2.03%", "elapsed_time": "18m 8s", "remaining_time": "14h 37m 23s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.045932} +{"loss": 0.42094727, "grad_norm": 0.55820203, "learning_rate": 7.33e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86202496, "epoch": 0.02229203, "global_step/max_steps": "55/2468", "percentage": "2.23%", "elapsed_time": "19m 15s", "remaining_time": "14h 5m 0s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.047593} +{"loss": 0.39628906, "grad_norm": 0.6299156, "learning_rate": 8e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86986344, "epoch": 0.02431857, "global_step/max_steps": "60/2468", "percentage": "2.43%", "elapsed_time": "20m 33s", "remaining_time": "13h 45m 12s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.048635} +{"loss": 0.36118164, "grad_norm": 0.32957265, "learning_rate": 8.67e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88104669, "epoch": 0.02634512, "global_step/max_steps": "65/2468", "percentage": "2.63%", "elapsed_time": "22m 8s", "remaining_time": "13h 38m 43s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.048917} +{"loss": 0.42099609, "grad_norm": 0.54532605, "learning_rate": 9.33e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86122752, "epoch": 0.02837167, "global_step/max_steps": "70/2468", "percentage": "2.84%", "elapsed_time": "23m 18s", "remaining_time": "13h 18m 26s", "memory(GiB)": 43.79, "train_speed(iter/s)": 0.050056} +{"loss": 0.36569824, "grad_norm": 0.35173345, "learning_rate": 1e-05, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88242273, "epoch": 0.03039822, "global_step/max_steps": "75/2468", "percentage": "3.04%", "elapsed_time": "24m 28s", "remaining_time": "13h 1m 7s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.051059} +{"loss": 0.40522461, "grad_norm": 0.63264841, "learning_rate": 1e-05, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86698865, "epoch": 0.03242476, "global_step/max_steps": "80/2468", "percentage": "3.24%", "elapsed_time": "25m 36s", "remaining_time": "12h 44m 26s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.052065} +{"loss": 0.37475586, "grad_norm": 0.41709852, "learning_rate": 1e-05, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88178928, "epoch": 0.03445131, "global_step/max_steps": "85/2468", "percentage": "3.44%", "elapsed_time": "26m 47s", "remaining_time": "12h 30m 56s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.05289} +{"loss": 0.39758301, "grad_norm": 0.63641804, "learning_rate": 1e-05, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86866571, "epoch": 0.03647786, "global_step/max_steps": "90/2468", "percentage": "3.65%", "elapsed_time": "27m 57s", "remaining_time": "12h 18m 43s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.053651} +{"loss": 0.38676758, "grad_norm": 0.42150235, "learning_rate": 1e-05, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87625037, "epoch": 0.03850441, "global_step/max_steps": "95/2468", "percentage": "3.85%", "elapsed_time": "29m 40s", "remaining_time": "12h 21m 20s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.05335} +{"loss": 0.39499512, "grad_norm": 0.55006528, "learning_rate": 1e-05, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87043539, "epoch": 0.04053096, "global_step/max_steps": "100/2468", "percentage": "4.05%", "elapsed_time": "30m 47s", "remaining_time": "12h 9m 8s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.054127} +{"loss": 0.38718262, "grad_norm": 0.41933143, "learning_rate": 1e-05, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87428904, "epoch": 0.0425575, "global_step/max_steps": "105/2468", "percentage": "4.25%", "elapsed_time": "31m 55s", "remaining_time": "11h 58m 38s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.054803} +{"loss": 0.39624023, "grad_norm": 0.47089764, "learning_rate": 9.99e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86786009, "epoch": 0.04458405, "global_step/max_steps": "110/2468", "percentage": "4.46%", "elapsed_time": "33m 6s", "remaining_time": "11h 49m 42s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.055375} +{"loss": 0.39257812, "grad_norm": 0.45688829, "learning_rate": 9.99e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86930565, "epoch": 0.0466106, "global_step/max_steps": "115/2468", "percentage": "4.66%", "elapsed_time": "34m 16s", "remaining_time": "11h 41m 18s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.05592} +{"loss": 0.37563477, "grad_norm": 0.4688974, "learning_rate": 9.99e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8766478, "epoch": 0.04863715, "global_step/max_steps": "120/2468", "percentage": "4.86%", "elapsed_time": "35m 25s", "remaining_time": "11h 33m 15s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.056448} +{"loss": 0.37924805, "grad_norm": 0.42888066, "learning_rate": 9.99e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87597559, "epoch": 0.05066369, "global_step/max_steps": "125/2468", "percentage": "5.06%", "elapsed_time": "36m 37s", "remaining_time": "11h 26m 21s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.056894} +{"loss": 0.42316895, "grad_norm": 0.46003604, "learning_rate": 9.99e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86165428, "epoch": 0.05269024, "global_step/max_steps": "130/2468", "percentage": "5.27%", "elapsed_time": "37m 44s", "remaining_time": "11h 18m 53s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.057398} +{"loss": 0.36162109, "grad_norm": 0.53018314, "learning_rate": 9.98e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87863799, "epoch": 0.05471679, "global_step/max_steps": "135/2468", "percentage": "5.47%", "elapsed_time": "38m 56s", "remaining_time": "11h 12m 51s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.057788} +{"loss": 0.40427246, "grad_norm": 0.47949114, "learning_rate": 9.98e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86586581, "epoch": 0.05674334, "global_step/max_steps": "140/2468", "percentage": "5.67%", "elapsed_time": "40m 7s", "remaining_time": "11h 7m 7s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.058159} +{"loss": 0.36850586, "grad_norm": 0.48405263, "learning_rate": 9.98e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87755759, "epoch": 0.05876989, "global_step/max_steps": "145/2468", "percentage": "5.88%", "elapsed_time": "41m 23s", "remaining_time": "11h 3m 10s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.058381} +{"loss": 0.37258301, "grad_norm": 0.40476394, "learning_rate": 9.98e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87729315, "epoch": 0.06079643, "global_step/max_steps": "150/2468", "percentage": "6.08%", "elapsed_time": "42m 36s", "remaining_time": "10h 58m 20s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.058683} +{"loss": 0.37080078, "grad_norm": 0.42546639, "learning_rate": 9.97e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88020198, "epoch": 0.06282298, "global_step/max_steps": "155/2468", "percentage": "6.28%", "elapsed_time": "43m 50s", "remaining_time": "10h 54m 14s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.058924} +{"loss": 0.40778809, "grad_norm": 0.56003207, "learning_rate": 9.97e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86620456, "epoch": 0.06484953, "global_step/max_steps": "160/2468", "percentage": "6.48%", "elapsed_time": "45m 36s", "remaining_time": "10h 57m 48s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.058477} +{"loss": 0.3838623, "grad_norm": 0.37908798, "learning_rate": 9.97e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87286472, "epoch": 0.06687608, "global_step/max_steps": "165/2468", "percentage": "6.69%", "elapsed_time": "46m 42s", "remaining_time": "10h 51m 57s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.058873} +{"loss": 0.42102051, "grad_norm": 0.49446025, "learning_rate": 9.96e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86088958, "epoch": 0.06890262, "global_step/max_steps": "170/2468", "percentage": "6.89%", "elapsed_time": "47m 50s", "remaining_time": "10h 46m 37s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.05923} +{"loss": 0.37773438, "grad_norm": 0.48118705, "learning_rate": 9.96e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87578978, "epoch": 0.07092917, "global_step/max_steps": "175/2468", "percentage": "7.09%", "elapsed_time": "48m 59s", "remaining_time": "10h 41m 54s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.059536} +{"loss": 0.41367188, "grad_norm": 0.5156424, "learning_rate": 9.95e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86440254, "epoch": 0.07295572, "global_step/max_steps": "180/2468", "percentage": "7.29%", "elapsed_time": "50m 9s", "remaining_time": "10h 37m 28s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.05982} +{"loss": 0.38076172, "grad_norm": 0.37811586, "learning_rate": 9.95e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87481126, "epoch": 0.07498227, "global_step/max_steps": "185/2468", "percentage": "7.50%", "elapsed_time": "51m 18s", "remaining_time": "10h 33m 16s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.060085} +{"loss": 0.39353027, "grad_norm": 0.42654335, "learning_rate": 9.94e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87205479, "epoch": 0.07700882, "global_step/max_steps": "190/2468", "percentage": "7.70%", "elapsed_time": "52m 31s", "remaining_time": "10h 29m 41s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.060294} +{"loss": 0.39211426, "grad_norm": 0.43498939, "learning_rate": 9.94e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87064061, "epoch": 0.07903536, "global_step/max_steps": "195/2468", "percentage": "7.90%", "elapsed_time": "53m 38s", "remaining_time": "10h 25m 15s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.060588} +{"loss": 0.37509766, "grad_norm": 0.43034092, "learning_rate": 9.93e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87616691, "epoch": 0.08106191, "global_step/max_steps": "200/2468", "percentage": "8.10%", "elapsed_time": "54m 58s", "remaining_time": "10h 23m 25s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.060633} +{"loss": 0.39194336, "grad_norm": 0.55232894, "learning_rate": 9.93e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87044231, "epoch": 0.08308846, "global_step/max_steps": "205/2468", "percentage": "8.31%", "elapsed_time": "56m 31s", "remaining_time": "10h 24m 0s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.060442} +{"loss": 0.39956055, "grad_norm": 0.49434412, "learning_rate": 9.92e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8697049, "epoch": 0.08511501, "global_step/max_steps": "210/2468", "percentage": "8.51%", "elapsed_time": "57m 40s", "remaining_time": "10h 20m 10s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.060681} +{"loss": 0.41660156, "grad_norm": 0.50669742, "learning_rate": 9.92e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86149231, "epoch": 0.08714155, "global_step/max_steps": "215/2468", "percentage": "8.71%", "elapsed_time": "58m 48s", "remaining_time": "10h 16m 20s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.060924} +{"loss": 0.39150391, "grad_norm": 0.43591195, "learning_rate": 9.91e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87271037, "epoch": 0.0891681, "global_step/max_steps": "220/2468", "percentage": "8.91%", "elapsed_time": "59m 56s", "remaining_time": "10h 12m 26s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.061176} +{"loss": 0.38364258, "grad_norm": 0.50964743, "learning_rate": 9.9e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87322935, "epoch": 0.09119465, "global_step/max_steps": "225/2468", "percentage": "9.12%", "elapsed_time": "1h 1m 4s", "remaining_time": "10h 8m 50s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.0614} +{"loss": 0.42202148, "grad_norm": 0.41969901, "learning_rate": 9.9e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86240105, "epoch": 0.0932212, "global_step/max_steps": "230/2468", "percentage": "9.32%", "elapsed_time": "1h 2m 12s", "remaining_time": "10h 5m 15s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.061627} +{"loss": 0.38181152, "grad_norm": 0.52181619, "learning_rate": 9.89e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87515342, "epoch": 0.09524775, "global_step/max_steps": "235/2468", "percentage": "9.52%", "elapsed_time": "1h 3m 21s", "remaining_time": "10h 2m 6s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.061811} +{"loss": 0.39291992, "grad_norm": 0.37742889, "learning_rate": 9.88e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87003558, "epoch": 0.09727429, "global_step/max_steps": "240/2468", "percentage": "9.72%", "elapsed_time": "1h 4m 31s", "remaining_time": "9h 59m 3s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.061987} +{"loss": 0.39609375, "grad_norm": 0.46909821, "learning_rate": 9.88e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86885589, "epoch": 0.09930084, "global_step/max_steps": "245/2468", "percentage": "9.93%", "elapsed_time": "1h 5m 42s", "remaining_time": "9h 56m 9s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.062148} +{"loss": 0.37536621, "grad_norm": 0.64840215, "learning_rate": 9.87e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87631792, "epoch": 0.10132739, "global_step/max_steps": "250/2468", "percentage": "10.13%", "elapsed_time": "1h 6m 52s", "remaining_time": "9h 53m 18s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.062305} +{"loss": 0.38937988, "grad_norm": 0.51854694, "learning_rate": 9.86e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8709359, "epoch": 0.10335394, "global_step/max_steps": "255/2468", "percentage": "10.33%", "elapsed_time": "1h 8m 3s", "remaining_time": "9h 50m 37s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.062448} +{"loss": 0.40112305, "grad_norm": 0.48241207, "learning_rate": 9.85e-06, "tIoU": 0.99698467, "tRecall_30": 0.9979166, "tRecall_50": 0.9979166, "tRecall_70": 0.9979166, "token_acc": 0.86794682, "epoch": 0.10538048, "global_step/max_steps": "260/2468", "percentage": "10.53%", "elapsed_time": "1h 9m 12s", "remaining_time": "9h 47m 45s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.062611} +{"loss": 0.40214844, "grad_norm": 0.49434677, "learning_rate": 9.85e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8677979, "epoch": 0.10740703, "global_step/max_steps": "265/2468", "percentage": "10.74%", "elapsed_time": "1h 10m 57s", "remaining_time": "9h 49m 53s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.062243} +{"loss": 0.3967041, "grad_norm": 0.5944103, "learning_rate": 9.84e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87122271, "epoch": 0.10943358, "global_step/max_steps": "270/2468", "percentage": "10.94%", "elapsed_time": "1h 12m 3s", "remaining_time": "9h 46m 36s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.062449} +{"loss": 0.36860352, "grad_norm": 0.45184132, "learning_rate": 9.83e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8792639, "epoch": 0.11146013, "global_step/max_steps": "275/2468", "percentage": "11.14%", "elapsed_time": "1h 13m 12s", "remaining_time": "9h 43m 49s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.062605} +{"loss": 0.40834961, "grad_norm": 0.49861455, "learning_rate": 9.82e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86567164, "epoch": 0.11348668, "global_step/max_steps": "280/2468", "percentage": "11.35%", "elapsed_time": "1h 14m 20s", "remaining_time": "9h 40m 54s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.062776} +{"loss": 0.39873047, "grad_norm": 0.41846049, "learning_rate": 9.81e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86864179, "epoch": 0.11551322, "global_step/max_steps": "285/2468", "percentage": "11.55%", "elapsed_time": "1h 15m 26s", "remaining_time": "9h 37m 49s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.062966} +{"loss": 0.40766602, "grad_norm": 0.57323301, "learning_rate": 9.8e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86541678, "epoch": 0.11753977, "global_step/max_steps": "290/2468", "percentage": "11.75%", "elapsed_time": "1h 16m 32s", "remaining_time": "9h 34m 48s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.063151} +{"loss": 0.39223633, "grad_norm": 0.42870671, "learning_rate": 9.79e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87069782, "epoch": 0.11956632, "global_step/max_steps": "295/2468", "percentage": "11.95%", "elapsed_time": "1h 17m 41s", "remaining_time": "9h 32m 15s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.063286} +{"loss": 0.41762695, "grad_norm": 0.78023779, "learning_rate": 9.78e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8623497, "epoch": 0.12159287, "global_step/max_steps": "300/2468", "percentage": "12.16%", "elapsed_time": "1h 18m 48s", "remaining_time": "9h 29m 28s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.063451} +{"loss": 0.38930664, "grad_norm": 0.55232304, "learning_rate": 9.77e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86975858, "epoch": 0.12361941, "global_step/max_steps": "305/2468", "percentage": "12.36%", "elapsed_time": "1h 19m 55s", "remaining_time": "9h 26m 51s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.063596} +{"loss": 0.40200195, "grad_norm": 0.48368698, "learning_rate": 9.76e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86823667, "epoch": 0.12564596, "global_step/max_steps": "310/2468", "percentage": "12.56%", "elapsed_time": "1h 21m 5s", "remaining_time": "9h 24m 27s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.063718} +{"loss": 0.37734375, "grad_norm": 0.33764985, "learning_rate": 9.75e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87675544, "epoch": 0.12767251, "global_step/max_steps": "315/2468", "percentage": "12.76%", "elapsed_time": "1h 22m 16s", "remaining_time": "9h 22m 22s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.063807} +{"loss": 0.37670898, "grad_norm": 0.26270628, "learning_rate": 9.74e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87747235, "epoch": 0.12969906, "global_step/max_steps": "320/2468", "percentage": "12.97%", "elapsed_time": "1h 23m 25s", "remaining_time": "9h 20m 0s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.063928} +{"loss": 0.38178711, "grad_norm": 0.39964983, "learning_rate": 9.73e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87234108, "epoch": 0.13172561, "global_step/max_steps": "325/2468", "percentage": "13.17%", "elapsed_time": "1h 24m 36s", "remaining_time": "9h 17m 53s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.064021} +{"loss": 0.35991211, "grad_norm": 0.45583984, "learning_rate": 9.72e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8835973, "epoch": 0.13375215, "global_step/max_steps": "330/2468", "percentage": "13.37%", "elapsed_time": "1h 26m 6s", "remaining_time": "9h 17m 54s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.063869} +{"loss": 0.41782227, "grad_norm": 0.63836116, "learning_rate": 9.71e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86180288, "epoch": 0.1357787, "global_step/max_steps": "335/2468", "percentage": "13.57%", "elapsed_time": "1h 27m 14s", "remaining_time": "9h 15m 26s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.064003} +{"loss": 0.41381836, "grad_norm": 0.61638314, "learning_rate": 9.7e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8630408, "epoch": 0.13780525, "global_step/max_steps": "340/2468", "percentage": "13.78%", "elapsed_time": "1h 28m 23s", "remaining_time": "9h 13m 12s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.064111} +{"loss": 0.40109863, "grad_norm": 0.5585286, "learning_rate": 9.69e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86683679, "epoch": 0.1398318, "global_step/max_steps": "345/2468", "percentage": "13.98%", "elapsed_time": "1h 29m 30s", "remaining_time": "9h 10m 50s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.064236} +{"loss": 0.40559082, "grad_norm": 0.53249776, "learning_rate": 9.68e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86569147, "epoch": 0.14185834, "global_step/max_steps": "350/2468", "percentage": "14.18%", "elapsed_time": "1h 30m 38s", "remaining_time": "9h 8m 32s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.064353} +{"loss": 0.37319336, "grad_norm": 0.52015889, "learning_rate": 9.67e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87625515, "epoch": 0.14388489, "global_step/max_steps": "355/2468", "percentage": "14.38%", "elapsed_time": "1h 31m 47s", "remaining_time": "9h 6m 23s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.064454} +{"loss": 0.39013672, "grad_norm": 0.39076343, "learning_rate": 9.65e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87100043, "epoch": 0.14591144, "global_step/max_steps": "360/2468", "percentage": "14.59%", "elapsed_time": "1h 32m 54s", "remaining_time": "9h 3m 59s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.064583} +{"loss": 0.38049316, "grad_norm": 0.46754771, "learning_rate": 9.64e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87503316, "epoch": 0.14793799, "global_step/max_steps": "365/2468", "percentage": "14.79%", "elapsed_time": "1h 34m 6s", "remaining_time": "9h 2m 14s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.064639} +{"loss": 0.3907959, "grad_norm": 0.51553816, "learning_rate": 9.63e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86919688, "epoch": 0.14996454, "global_step/max_steps": "370/2468", "percentage": "14.99%", "elapsed_time": "1h 35m 13s", "remaining_time": "8h 59m 59s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.064754} +{"loss": 0.40219727, "grad_norm": 0.45304245, "learning_rate": 9.62e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86732403, "epoch": 0.15199108, "global_step/max_steps": "375/2468", "percentage": "15.19%", "elapsed_time": "1h 36m 21s", "remaining_time": "8h 57m 48s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.064863} +{"loss": 0.39306641, "grad_norm": 0.46315023, "learning_rate": 9.6e-06, "tIoU": 0.99996452, "tRecall_30": 1.0, "tRecall_50": 0.99980469, "tRecall_70": 0.99980469, "token_acc": 0.86920226, "epoch": 0.15401763, "global_step/max_steps": "380/2468", "percentage": "15.40%", "elapsed_time": "1h 37m 27s", "remaining_time": "8h 55m 28s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.064989} +{"loss": 0.38962402, "grad_norm": 0.45243508, "learning_rate": 9.59e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87199905, "epoch": 0.15604418, "global_step/max_steps": "385/2468", "percentage": "15.60%", "elapsed_time": "1h 38m 35s", "remaining_time": "8h 53m 27s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065079} +{"loss": 0.37585449, "grad_norm": 0.30289549, "learning_rate": 9.58e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88077843, "epoch": 0.15807073, "global_step/max_steps": "390/2468", "percentage": "15.80%", "elapsed_time": "1h 39m 47s", "remaining_time": "8h 51m 40s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.06514} +{"loss": 0.36181641, "grad_norm": 0.42049021, "learning_rate": 9.57e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88223344, "epoch": 0.16009727, "global_step/max_steps": "395/2468", "percentage": "16.00%", "elapsed_time": "1h 40m 58s", "remaining_time": "8h 49m 53s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065201} +{"loss": 0.3998291, "grad_norm": 0.54576653, "learning_rate": 9.55e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86883312, "epoch": 0.16212382, "global_step/max_steps": "400/2468", "percentage": "16.21%", "elapsed_time": "1h 42m 7s", "remaining_time": "8h 47m 58s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.06528} +{"loss": 0.40454102, "grad_norm": 0.59216869, "learning_rate": 9.54e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86737379, "epoch": 0.16415037, "global_step/max_steps": "405/2468", "percentage": "16.41%", "elapsed_time": "1h 43m 33s", "remaining_time": "8h 47m 31s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065179} +{"loss": 0.40275879, "grad_norm": 0.44998884, "learning_rate": 9.52e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86596217, "epoch": 0.16617692, "global_step/max_steps": "410/2468", "percentage": "16.61%", "elapsed_time": "1h 44m 41s", "remaining_time": "8h 45m 28s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065273} +{"loss": 0.39575195, "grad_norm": 0.47718403, "learning_rate": 9.51e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86840103, "epoch": 0.16820347, "global_step/max_steps": "415/2468", "percentage": "16.82%", "elapsed_time": "1h 45m 49s", "remaining_time": "8h 43m 29s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065362} +{"loss": 0.3904541, "grad_norm": 0.46238291, "learning_rate": 9.5e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86884246, "epoch": 0.17023001, "global_step/max_steps": "420/2468", "percentage": "17.02%", "elapsed_time": "1h 47m 1s", "remaining_time": "8h 41m 53s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065404} +{"loss": 0.40148926, "grad_norm": 0.38424453, "learning_rate": 9.48e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86842986, "epoch": 0.17225656, "global_step/max_steps": "425/2468", "percentage": "17.22%", "elapsed_time": "1h 48m 9s", "remaining_time": "8h 39m 53s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065495} +{"loss": 0.38728027, "grad_norm": 0.46710488, "learning_rate": 9.47e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87241539, "epoch": 0.17428311, "global_step/max_steps": "430/2468", "percentage": "17.42%", "elapsed_time": "1h 49m 18s", "remaining_time": "8h 38m 3s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065566} +{"loss": 0.38771973, "grad_norm": 0.51993591, "learning_rate": 9.45e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87418142, "epoch": 0.17630966, "global_step/max_steps": "435/2468", "percentage": "17.63%", "elapsed_time": "1h 50m 25s", "remaining_time": "8h 36m 6s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065651} +{"loss": 0.3878418, "grad_norm": 0.42699197, "learning_rate": 9.44e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87102288, "epoch": 0.1783362, "global_step/max_steps": "440/2468", "percentage": "17.83%", "elapsed_time": "1h 51m 39s", "remaining_time": "8h 34m 39s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065674} +{"loss": 0.37880859, "grad_norm": 0.49447265, "learning_rate": 9.42e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87539885, "epoch": 0.18036275, "global_step/max_steps": "445/2468", "percentage": "18.03%", "elapsed_time": "1h 52m 47s", "remaining_time": "8h 32m 46s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065754} +{"loss": 0.36621094, "grad_norm": 0.52266312, "learning_rate": 9.41e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88128858, "epoch": 0.1823893, "global_step/max_steps": "450/2468", "percentage": "18.23%", "elapsed_time": "1h 53m 57s", "remaining_time": "8h 31m 3s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065811} +{"loss": 0.39199219, "grad_norm": 0.4460074, "learning_rate": 9.39e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86871198, "epoch": 0.18441585, "global_step/max_steps": "455/2468", "percentage": "18.44%", "elapsed_time": "1h 55m 4s", "remaining_time": "8h 29m 5s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065901} +{"loss": 0.39545898, "grad_norm": 0.41100502, "learning_rate": 9.37e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86945722, "epoch": 0.1864424, "global_step/max_steps": "460/2468", "percentage": "18.64%", "elapsed_time": "1h 56m 13s", "remaining_time": "8h 27m 20s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.065965} +{"loss": 0.3611084, "grad_norm": 0.43790925, "learning_rate": 9.36e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88465213, "epoch": 0.18846894, "global_step/max_steps": "465/2468", "percentage": "18.84%", "elapsed_time": "1h 57m 23s", "remaining_time": "8h 25m 41s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066014} +{"loss": 0.38503418, "grad_norm": 0.44727439, "learning_rate": 9.34e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87307635, "epoch": 0.19049549, "global_step/max_steps": "470/2468", "percentage": "19.04%", "elapsed_time": "1h 58m 31s", "remaining_time": "8h 23m 53s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066087} +{"loss": 0.39831543, "grad_norm": 0.49198452, "learning_rate": 9.33e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86712223, "epoch": 0.19252204, "global_step/max_steps": "475/2468", "percentage": "19.25%", "elapsed_time": "1h 59m 39s", "remaining_time": "8h 22m 4s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066158} +{"loss": 0.37763672, "grad_norm": 0.42646599, "learning_rate": 9.31e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8753093, "epoch": 0.19454859, "global_step/max_steps": "480/2468", "percentage": "19.45%", "elapsed_time": "2h 0m 49s", "remaining_time": "8h 20m 25s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.06621} +{"loss": 0.40056152, "grad_norm": 0.43917012, "learning_rate": 9.29e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86701072, "epoch": 0.19657513, "global_step/max_steps": "485/2468", "percentage": "19.65%", "elapsed_time": "2h 1m 57s", "remaining_time": "8h 18m 37s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066282} +{"loss": 0.39455566, "grad_norm": 0.3988682, "learning_rate": 9.28e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87100313, "epoch": 0.19860168, "global_step/max_steps": "490/2468", "percentage": "19.85%", "elapsed_time": "2h 3m 5s", "remaining_time": "8h 16m 52s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066348} +{"loss": 0.37817383, "grad_norm": 0.45519567, "learning_rate": 9.26e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87443375, "epoch": 0.20062823, "global_step/max_steps": "495/2468", "percentage": "20.06%", "elapsed_time": "2h 4m 13s", "remaining_time": "8h 15m 8s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066413} +{"loss": 0.34606934, "grad_norm": 0.43706313, "learning_rate": 9.24e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88400014, "epoch": 0.20265478, "global_step/max_steps": "500/2468", "percentage": "20.26%", "elapsed_time": "2h 5m 22s", "remaining_time": "8h 13m 29s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066465} +{"loss": 0.36459961, "grad_norm": 0.39965802, "learning_rate": 9.22e-06, "tIoU": 0.99986496, "tRecall_30": 0.99985685, "tRecall_50": 0.99985685, "tRecall_70": 0.99985685, "token_acc": 0.8795211, "epoch": 0.20468133, "global_step/max_steps": "505/2468", "percentage": "20.46%", "elapsed_time": "2h 6m 31s", "remaining_time": "8h 11m 49s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066522} +{"loss": 0.38718262, "grad_norm": 0.48396146, "learning_rate": 9.21e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87238624, "epoch": 0.20670787, "global_step/max_steps": "510/2468", "percentage": "20.66%", "elapsed_time": "2h 7m 40s", "remaining_time": "8h 10m 9s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066576} +{"loss": 0.3907959, "grad_norm": 0.56390071, "learning_rate": 9.19e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87024648, "epoch": 0.20873442, "global_step/max_steps": "515/2468", "percentage": "20.87%", "elapsed_time": "2h 8m 52s", "remaining_time": "8h 8m 42s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066605} +{"loss": 0.37683105, "grad_norm": 0.46063733, "learning_rate": 9.17e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87671459, "epoch": 0.21076097, "global_step/max_steps": "520/2468", "percentage": "21.07%", "elapsed_time": "2h 10m 1s", "remaining_time": "8h 7m 4s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066655} +{"loss": 0.37785645, "grad_norm": 0.41434205, "learning_rate": 9.15e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87482133, "epoch": 0.21278752, "global_step/max_steps": "525/2468", "percentage": "21.27%", "elapsed_time": "2h 11m 13s", "remaining_time": "8h 5m 37s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066683} +{"loss": 0.40649414, "grad_norm": 0.56855142, "learning_rate": 9.13e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86428337, "epoch": 0.21481406, "global_step/max_steps": "530/2468", "percentage": "21.47%", "elapsed_time": "2h 12m 21s", "remaining_time": "8h 3m 57s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066742} +{"loss": 0.38723145, "grad_norm": 0.54202831, "learning_rate": 9.12e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87243784, "epoch": 0.21684061, "global_step/max_steps": "535/2468", "percentage": "21.68%", "elapsed_time": "2h 13m 42s", "remaining_time": "8h 3m 6s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066686} +{"loss": 0.41130371, "grad_norm": 0.64008129, "learning_rate": 9.1e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87053283, "epoch": 0.21886716, "global_step/max_steps": "540/2468", "percentage": "21.88%", "elapsed_time": "2h 14m 51s", "remaining_time": "8h 1m 31s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066734} +{"loss": 0.40126953, "grad_norm": 0.6898703, "learning_rate": 9.08e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86718375, "epoch": 0.22089371, "global_step/max_steps": "545/2468", "percentage": "22.08%", "elapsed_time": "2h 16m 1s", "remaining_time": "7h 59m 56s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066779} +{"loss": 0.35925293, "grad_norm": 0.44395372, "learning_rate": 9.06e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88155469, "epoch": 0.22292026, "global_step/max_steps": "550/2468", "percentage": "22.29%", "elapsed_time": "2h 17m 10s", "remaining_time": "7h 58m 21s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066825} +{"loss": 0.37060547, "grad_norm": 0.38134971, "learning_rate": 9.04e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87783136, "epoch": 0.2249468, "global_step/max_steps": "555/2468", "percentage": "22.49%", "elapsed_time": "2h 18m 19s", "remaining_time": "7h 56m 46s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066873} +{"loss": 0.40673828, "grad_norm": 0.35967746, "learning_rate": 9.02e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86522957, "epoch": 0.22697335, "global_step/max_steps": "560/2468", "percentage": "22.69%", "elapsed_time": "2h 19m 29s", "remaining_time": "7h 55m 17s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066907} +{"loss": 0.38093262, "grad_norm": 0.6109125, "learning_rate": 9e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8743921, "epoch": 0.2289999, "global_step/max_steps": "565/2468", "percentage": "22.89%", "elapsed_time": "2h 20m 40s", "remaining_time": "7h 53m 47s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066943} +{"loss": 0.38361816, "grad_norm": 0.4746131, "learning_rate": 8.98e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87451355, "epoch": 0.23102645, "global_step/max_steps": "570/2468", "percentage": "23.10%", "elapsed_time": "2h 21m 50s", "remaining_time": "7h 52m 18s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.066976} +{"loss": 0.34702148, "grad_norm": 0.410402, "learning_rate": 8.96e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88637917, "epoch": 0.23305299, "global_step/max_steps": "575/2468", "percentage": "23.30%", "elapsed_time": "2h 23m 1s", "remaining_time": "7h 50m 52s", "memory(GiB)": 54.18, "train_speed(iter/s)": 0.067002} +{"loss": 0.37160645, "grad_norm": 0.42033172, "learning_rate": 8.94e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88029468, "epoch": 0.23507954, "global_step/max_steps": "580/2468", "percentage": "23.50%", "elapsed_time": "2h 24m 13s", "remaining_time": "7h 49m 28s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067024} +{"loss": 0.39038086, "grad_norm": 0.44602793, "learning_rate": 8.92e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87202341, "epoch": 0.23710609, "global_step/max_steps": "585/2468", "percentage": "23.70%", "elapsed_time": "2h 25m 22s", "remaining_time": "7h 47m 57s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067064} +{"loss": 0.4015625, "grad_norm": 0.93072116, "learning_rate": 8.9e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86594621, "epoch": 0.23913264, "global_step/max_steps": "590/2468", "percentage": "23.91%", "elapsed_time": "2h 26m 27s", "remaining_time": "7h 46m 10s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067142} +{"loss": 0.39284668, "grad_norm": 0.44321755, "learning_rate": 8.88e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8695431, "epoch": 0.24115919, "global_step/max_steps": "595/2468", "percentage": "24.11%", "elapsed_time": "2h 27m 35s", "remaining_time": "7h 44m 35s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067192} +{"loss": 0.37045898, "grad_norm": 0.37842602, "learning_rate": 8.86e-06, "tIoU": 0.99988508, "tRecall_30": 0.99989748, "tRecall_50": 0.99989748, "tRecall_70": 0.99989748, "token_acc": 0.87855289, "epoch": 0.24318573, "global_step/max_steps": "600/2468", "percentage": "24.31%", "elapsed_time": "2h 28m 44s", "remaining_time": "7h 43m 4s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067233} +{"loss": 0.37653809, "grad_norm": 0.45076823, "learning_rate": 8.84e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87669754, "epoch": 0.24521228, "global_step/max_steps": "605/2468", "percentage": "24.51%", "elapsed_time": "2h 29m 51s", "remaining_time": "7h 41m 28s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067285} +{"loss": 0.36555176, "grad_norm": 0.48391059, "learning_rate": 8.82e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88251254, "epoch": 0.24723883, "global_step/max_steps": "610/2468", "percentage": "24.72%", "elapsed_time": "2h 31m 2s", "remaining_time": "7h 40m 2s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067313} +{"loss": 0.37539062, "grad_norm": 0.93918037, "learning_rate": 8.8e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87493147, "epoch": 0.24926538, "global_step/max_steps": "615/2468", "percentage": "24.92%", "elapsed_time": "2h 32m 11s", "remaining_time": "7h 38m 33s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067349} +{"loss": 0.40258789, "grad_norm": 0.53651488, "learning_rate": 8.77e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8673238, "epoch": 0.25129192, "global_step/max_steps": "620/2468", "percentage": "25.12%", "elapsed_time": "2h 33m 28s", "remaining_time": "7h 37m 26s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067331} +{"loss": 0.36464844, "grad_norm": 0.41079307, "learning_rate": 8.75e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88035347, "epoch": 0.25331847, "global_step/max_steps": "625/2468", "percentage": "25.32%", "elapsed_time": "2h 34m 38s", "remaining_time": "7h 35m 59s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067362} +{"loss": 0.40307617, "grad_norm": 0.36408988, "learning_rate": 8.73e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86757919, "epoch": 0.25534502, "global_step/max_steps": "630/2468", "percentage": "25.53%", "elapsed_time": "2h 35m 47s", "remaining_time": "7h 34m 31s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067397} +{"loss": 0.37197266, "grad_norm": 0.40313908, "learning_rate": 8.71e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87791348, "epoch": 0.25737157, "global_step/max_steps": "635/2468", "percentage": "25.73%", "elapsed_time": "2h 36m 58s", "remaining_time": "7h 33m 8s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067418} +{"loss": 0.38249512, "grad_norm": 0.48640084, "learning_rate": 8.69e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87330074, "epoch": 0.25939812, "global_step/max_steps": "640/2468", "percentage": "25.93%", "elapsed_time": "2h 38m 12s", "remaining_time": "7h 31m 52s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067423} +{"loss": 0.38925781, "grad_norm": 0.44566435, "learning_rate": 8.66e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87049249, "epoch": 0.26142466, "global_step/max_steps": "645/2468", "percentage": "26.13%", "elapsed_time": "2h 39m 21s", "remaining_time": "7h 30m 23s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067461} +{"loss": 0.38217773, "grad_norm": 0.37806481, "learning_rate": 8.64e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87410216, "epoch": 0.26345121, "global_step/max_steps": "650/2468", "percentage": "26.34%", "elapsed_time": "2h 40m 28s", "remaining_time": "7h 28m 51s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067506} +{"loss": 0.37209473, "grad_norm": 0.44057775, "learning_rate": 8.62e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87780303, "epoch": 0.26547776, "global_step/max_steps": "655/2468", "percentage": "26.54%", "elapsed_time": "2h 41m 39s", "remaining_time": "7h 27m 26s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067531} +{"loss": 0.3715332, "grad_norm": 0.47154233, "learning_rate": 8.6e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88014892, "epoch": 0.26750431, "global_step/max_steps": "660/2468", "percentage": "26.74%", "elapsed_time": "2h 42m 51s", "remaining_time": "7h 26m 8s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067543} +{"loss": 0.37375488, "grad_norm": 0.5181433, "learning_rate": 8.57e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87510642, "epoch": 0.26953085, "global_step/max_steps": "665/2468", "percentage": "26.94%", "elapsed_time": "2h 44m 1s", "remaining_time": "7h 24m 41s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067574} +{"loss": 0.36586914, "grad_norm": 0.41301727, "learning_rate": 8.55e-06, "tIoU": 0.99995232, "tRecall_30": 0.9999527, "tRecall_50": 0.9999527, "tRecall_70": 0.9999527, "token_acc": 0.87865423, "epoch": 0.2715574, "global_step/max_steps": "670/2468", "percentage": "27.15%", "elapsed_time": "2h 45m 12s", "remaining_time": "7h 23m 21s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067589} +{"loss": 0.37478027, "grad_norm": 0.47343948, "learning_rate": 8.53e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87629457, "epoch": 0.27358395, "global_step/max_steps": "675/2468", "percentage": "27.35%", "elapsed_time": "2h 46m 24s", "remaining_time": "7h 22m 0s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067608} +{"loss": 0.38325195, "grad_norm": 0.42571053, "learning_rate": 8.5e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87269584, "epoch": 0.2756105, "global_step/max_steps": "680/2468", "percentage": "27.55%", "elapsed_time": "2h 47m 32s", "remaining_time": "7h 20m 32s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067645} +{"loss": 0.40424805, "grad_norm": 0.47510353, "learning_rate": 8.48e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8653982, "epoch": 0.27763705, "global_step/max_steps": "685/2468", "percentage": "27.76%", "elapsed_time": "2h 48m 41s", "remaining_time": "7h 19m 5s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067677} +{"loss": 0.38100586, "grad_norm": 0.44467607, "learning_rate": 8.46e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87265812, "epoch": 0.27966359, "global_step/max_steps": "690/2468", "percentage": "27.96%", "elapsed_time": "2h 49m 48s", "remaining_time": "7h 17m 34s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067723} +{"loss": 0.39541016, "grad_norm": 0.48777613, "learning_rate": 8.43e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86818338, "epoch": 0.28169014, "global_step/max_steps": "695/2468", "percentage": "28.16%", "elapsed_time": "2h 50m 56s", "remaining_time": "7h 16m 4s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067765} +{"loss": 0.37702637, "grad_norm": 0.48598027, "learning_rate": 8.41e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87384822, "epoch": 0.28371669, "global_step/max_steps": "700/2468", "percentage": "28.36%", "elapsed_time": "2h 52m 6s", "remaining_time": "7h 14m 42s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067784} +{"loss": 0.35878906, "grad_norm": 0.33023939, "learning_rate": 8.39e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88111191, "epoch": 0.28574324, "global_step/max_steps": "705/2468", "percentage": "28.57%", "elapsed_time": "2h 53m 17s", "remaining_time": "7h 13m 21s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067803} +{"loss": 0.39621582, "grad_norm": 0.39953363, "learning_rate": 8.36e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86830079, "epoch": 0.28776978, "global_step/max_steps": "710/2468", "percentage": "28.77%", "elapsed_time": "2h 54m 24s", "remaining_time": "7h 11m 49s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067851} +{"loss": 0.37705078, "grad_norm": 0.2610229, "learning_rate": 8.34e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87590103, "epoch": 0.28979633, "global_step/max_steps": "715/2468", "percentage": "28.97%", "elapsed_time": "2h 55m 35s", "remaining_time": "7h 10m 30s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067866} +{"loss": 0.36098633, "grad_norm": 0.50307715, "learning_rate": 8.31e-06, "tIoU": 0.99960232, "tRecall_30": 0.99979172, "tRecall_50": 0.99979172, "tRecall_70": 0.99979172, "token_acc": 0.87926682, "epoch": 0.29182288, "global_step/max_steps": "720/2468", "percentage": "29.17%", "elapsed_time": "2h 56m 48s", "remaining_time": "7h 9m 15s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.06787} +{"loss": 0.34440918, "grad_norm": 0.41992182, "learning_rate": 8.29e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88508443, "epoch": 0.29384943, "global_step/max_steps": "725/2468", "percentage": "29.38%", "elapsed_time": "2h 57m 57s", "remaining_time": "7h 7m 50s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067899} +{"loss": 0.38041992, "grad_norm": 0.43417928, "learning_rate": 8.26e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87344118, "epoch": 0.29587598, "global_step/max_steps": "730/2468", "percentage": "29.58%", "elapsed_time": "2h 59m 5s", "remaining_time": "7h 6m 24s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067933} +{"loss": 0.35698242, "grad_norm": 0.40655169, "learning_rate": 8.24e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88513375, "epoch": 0.29790252, "global_step/max_steps": "735/2468", "percentage": "29.78%", "elapsed_time": "3h 0m 20s", "remaining_time": "7h 5m 11s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067929} +{"loss": 0.35112305, "grad_norm": 0.35124925, "learning_rate": 8.21e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88404901, "epoch": 0.29992907, "global_step/max_steps": "740/2468", "percentage": "29.98%", "elapsed_time": "3h 1m 37s", "remaining_time": "7h 4m 7s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067903} +{"loss": 0.4010498, "grad_norm": 0.37331146, "learning_rate": 8.19e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86763041, "epoch": 0.30195562, "global_step/max_steps": "745/2468", "percentage": "30.19%", "elapsed_time": "3h 2m 48s", "remaining_time": "7h 2m 48s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067919} +{"loss": 0.39763184, "grad_norm": 0.53629947, "learning_rate": 8.16e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86876919, "epoch": 0.30398217, "global_step/max_steps": "750/2468", "percentage": "30.39%", "elapsed_time": "3h 3m 56s", "remaining_time": "7h 1m 21s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067954} +{"loss": 0.37802734, "grad_norm": 0.71186101, "learning_rate": 8.14e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87428208, "epoch": 0.30600871, "global_step/max_steps": "755/2468", "percentage": "30.59%", "elapsed_time": "3h 5m 3s", "remaining_time": "6h 59m 52s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.067995} +{"loss": 0.38132324, "grad_norm": 0.4947682, "learning_rate": 8.11e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87581691, "epoch": 0.30803526, "global_step/max_steps": "760/2468", "percentage": "30.79%", "elapsed_time": "3h 6m 12s", "remaining_time": "6h 58m 29s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068021} +{"loss": 0.38432617, "grad_norm": 0.36296964, "learning_rate": 8.09e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87086887, "epoch": 0.31006181, "global_step/max_steps": "765/2468", "percentage": "31.00%", "elapsed_time": "3h 7m 19s", "remaining_time": "6h 57m 1s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068063} +{"loss": 0.35917969, "grad_norm": 0.41789612, "learning_rate": 8.06e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8794477, "epoch": 0.31208836, "global_step/max_steps": "770/2468", "percentage": "31.20%", "elapsed_time": "3h 8m 30s", "remaining_time": "6h 55m 41s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068079} +{"loss": 0.34238281, "grad_norm": 0.47259948, "learning_rate": 8.03e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88473674, "epoch": 0.31411491, "global_step/max_steps": "775/2468", "percentage": "31.40%", "elapsed_time": "3h 9m 40s", "remaining_time": "6h 54m 21s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068096} +{"loss": 0.38156738, "grad_norm": 0.36307207, "learning_rate": 8.01e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87374866, "epoch": 0.31614145, "global_step/max_steps": "780/2468", "percentage": "31.60%", "elapsed_time": "3h 10m 51s", "remaining_time": "6h 53m 2s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068113} +{"loss": 0.37084961, "grad_norm": 0.47526979, "learning_rate": 7.98e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87701502, "epoch": 0.318168, "global_step/max_steps": "785/2468", "percentage": "31.81%", "elapsed_time": "3h 12m 4s", "remaining_time": "6h 51m 47s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068117} +{"loss": 0.38996582, "grad_norm": 0.43980148, "learning_rate": 7.95e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8706937, "epoch": 0.32019455, "global_step/max_steps": "790/2468", "percentage": "32.01%", "elapsed_time": "3h 13m 11s", "remaining_time": "6h 50m 21s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068152} +{"loss": 0.33647461, "grad_norm": 0.48049763, "learning_rate": 7.93e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88571314, "epoch": 0.3222211, "global_step/max_steps": "795/2468", "percentage": "32.21%", "elapsed_time": "3h 14m 22s", "remaining_time": "6h 49m 1s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068169} +{"loss": 0.40803223, "grad_norm": 0.51272833, "learning_rate": 7.9e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86505762, "epoch": 0.32424764, "global_step/max_steps": "800/2468", "percentage": "32.41%", "elapsed_time": "3h 15m 28s", "remaining_time": "6h 47m 34s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068207} +{"loss": 0.39042969, "grad_norm": 0.56945974, "learning_rate": 7.87e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86940546, "epoch": 0.32627419, "global_step/max_steps": "805/2468", "percentage": "32.62%", "elapsed_time": "3h 16m 36s", "remaining_time": "6h 46m 9s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068241} +{"loss": 0.3904541, "grad_norm": 0.43124932, "learning_rate": 7.85e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87155571, "epoch": 0.32830074, "global_step/max_steps": "810/2468", "percentage": "32.82%", "elapsed_time": "3h 17m 44s", "remaining_time": "6h 44m 46s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.06827} +{"loss": 0.3909668, "grad_norm": 0.43032664, "learning_rate": 7.82e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86927959, "epoch": 0.33032729, "global_step/max_steps": "815/2468", "percentage": "33.02%", "elapsed_time": "3h 18m 51s", "remaining_time": "6h 43m 19s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068306} +{"loss": 0.35568848, "grad_norm": 0.34819663, "learning_rate": 7.79e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8809747, "epoch": 0.33235384, "global_step/max_steps": "820/2468", "percentage": "33.23%", "elapsed_time": "3h 20m 1s", "remaining_time": "6h 42m 1s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068322} +{"loss": 0.36413574, "grad_norm": 0.52338922, "learning_rate": 7.77e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87897258, "epoch": 0.33438038, "global_step/max_steps": "825/2468", "percentage": "33.43%", "elapsed_time": "3h 21m 9s", "remaining_time": "6h 40m 37s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068352} +{"loss": 0.35895996, "grad_norm": 0.43725345, "learning_rate": 7.74e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88460191, "epoch": 0.33640693, "global_step/max_steps": "830/2468", "percentage": "33.63%", "elapsed_time": "3h 22m 21s", "remaining_time": "6h 39m 21s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068358} +{"loss": 0.37680664, "grad_norm": 0.42524704, "learning_rate": 7.71e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87557729, "epoch": 0.33843348, "global_step/max_steps": "835/2468", "percentage": "33.83%", "elapsed_time": "3h 23m 29s", "remaining_time": "6h 37m 58s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068387} +{"loss": 0.33037109, "grad_norm": 0.28774598, "learning_rate": 7.68e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89222432, "epoch": 0.34046003, "global_step/max_steps": "840/2468", "percentage": "34.04%", "elapsed_time": "3h 24m 39s", "remaining_time": "6h 36m 39s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068404} +{"loss": 0.38400879, "grad_norm": 0.47681004, "learning_rate": 7.66e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87818597, "epoch": 0.34248657, "global_step/max_steps": "845/2468", "percentage": "34.24%", "elapsed_time": "3h 25m 48s", "remaining_time": "6h 35m 18s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068429} +{"loss": 0.36098633, "grad_norm": 0.3722887, "learning_rate": 7.63e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88048087, "epoch": 0.34451312, "global_step/max_steps": "850/2468", "percentage": "34.44%", "elapsed_time": "3h 26m 55s", "remaining_time": "6h 33m 53s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068462} +{"loss": 0.38364258, "grad_norm": 0.42258158, "learning_rate": 7.6e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87186077, "epoch": 0.34653967, "global_step/max_steps": "855/2468", "percentage": "34.64%", "elapsed_time": "3h 28m 2s", "remaining_time": "6h 32m 28s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068498} +{"loss": 0.40356445, "grad_norm": 0.50759369, "learning_rate": 7.57e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86658006, "epoch": 0.34856622, "global_step/max_steps": "860/2468", "percentage": "34.85%", "elapsed_time": "3h 29m 10s", "remaining_time": "6h 31m 5s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068525} +{"loss": 0.35847168, "grad_norm": 0.35151452, "learning_rate": 7.54e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88106062, "epoch": 0.35059277, "global_step/max_steps": "865/2468", "percentage": "35.05%", "elapsed_time": "3h 30m 20s", "remaining_time": "6h 29m 48s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068539} +{"loss": 0.36228027, "grad_norm": 0.44989818, "learning_rate": 7.52e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88611511, "epoch": 0.35261931, "global_step/max_steps": "870/2468", "percentage": "35.25%", "elapsed_time": "3h 31m 34s", "remaining_time": "6h 28m 36s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068535} +{"loss": 0.38457031, "grad_norm": 0.45056057, "learning_rate": 7.49e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87173183, "epoch": 0.35464586, "global_step/max_steps": "875/2468", "percentage": "35.45%", "elapsed_time": "3h 32m 40s", "remaining_time": "6h 27m 11s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.06857} +{"loss": 0.38085938, "grad_norm": 0.61112291, "learning_rate": 7.46e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87446016, "epoch": 0.35667241, "global_step/max_steps": "880/2468", "percentage": "35.66%", "elapsed_time": "3h 33m 48s", "remaining_time": "6h 25m 50s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068596} +{"loss": 0.40075684, "grad_norm": 0.54957247, "learning_rate": 7.43e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86892839, "epoch": 0.35869896, "global_step/max_steps": "885/2468", "percentage": "35.86%", "elapsed_time": "3h 34m 56s", "remaining_time": "6h 24m 28s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068623} +{"loss": 0.36074219, "grad_norm": 0.36970368, "learning_rate": 7.4e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87979352, "epoch": 0.3607255, "global_step/max_steps": "890/2468", "percentage": "36.06%", "elapsed_time": "3h 36m 7s", "remaining_time": "6h 23m 12s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068631} +{"loss": 0.3994873, "grad_norm": 0.50500458, "learning_rate": 7.37e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8684938, "epoch": 0.36275205, "global_step/max_steps": "895/2468", "percentage": "36.26%", "elapsed_time": "3h 37m 16s", "remaining_time": "6h 21m 52s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068653} +{"loss": 0.3815918, "grad_norm": 0.42693156, "learning_rate": 7.34e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87412498, "epoch": 0.3647786, "global_step/max_steps": "900/2468", "percentage": "36.47%", "elapsed_time": "3h 38m 23s", "remaining_time": "6h 20m 29s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068682} +{"loss": 0.37421875, "grad_norm": 0.42171878, "learning_rate": 7.31e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87871057, "epoch": 0.36680515, "global_step/max_steps": "905/2468", "percentage": "36.67%", "elapsed_time": "3h 39m 32s", "remaining_time": "6h 19m 9s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068705} +{"loss": 0.36838379, "grad_norm": 0.41110745, "learning_rate": 7.28e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87847635, "epoch": 0.3688317, "global_step/max_steps": "910/2468", "percentage": "36.87%", "elapsed_time": "3h 40m 41s", "remaining_time": "6h 17m 51s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068722} +{"loss": 0.35617676, "grad_norm": 0.43751952, "learning_rate": 7.26e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88099519, "epoch": 0.37085824, "global_step/max_steps": "915/2468", "percentage": "37.07%", "elapsed_time": "3h 41m 50s", "remaining_time": "6h 16m 31s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068743} +{"loss": 0.39624023, "grad_norm": 0.39494741, "learning_rate": 7.23e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86735761, "epoch": 0.37288479, "global_step/max_steps": "920/2468", "percentage": "37.28%", "elapsed_time": "3h 42m 57s", "remaining_time": "6h 15m 9s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.06877} +{"loss": 0.36325684, "grad_norm": 0.43488541, "learning_rate": 7.2e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87878581, "epoch": 0.37491134, "global_step/max_steps": "925/2468", "percentage": "37.48%", "elapsed_time": "3h 44m 6s", "remaining_time": "6h 13m 50s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068791} +{"loss": 0.37619629, "grad_norm": 0.45749152, "learning_rate": 7.17e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87565964, "epoch": 0.37693789, "global_step/max_steps": "930/2468", "percentage": "37.68%", "elapsed_time": "3h 45m 16s", "remaining_time": "6h 12m 33s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068803} +{"loss": 0.39418945, "grad_norm": 0.44833526, "learning_rate": 7.14e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86678204, "epoch": 0.37896443, "global_step/max_steps": "935/2468", "percentage": "37.88%", "elapsed_time": "3h 46m 23s", "remaining_time": "6h 11m 11s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068834} +{"loss": 0.34963379, "grad_norm": 0.52482146, "learning_rate": 7.11e-06, "tIoU": 0.99993896, "tRecall_30": 0.99992647, "tRecall_50": 0.99992647, "tRecall_70": 0.99992647, "token_acc": 0.8841317, "epoch": 0.38099098, "global_step/max_steps": "940/2468", "percentage": "38.09%", "elapsed_time": "3h 47m 33s", "remaining_time": "6h 9m 53s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068849} +{"loss": 0.37080078, "grad_norm": 0.39933234, "learning_rate": 7.08e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87464781, "epoch": 0.38301753, "global_step/max_steps": "945/2468", "percentage": "38.29%", "elapsed_time": "3h 48m 39s", "remaining_time": "6h 8m 30s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068882} +{"loss": 0.3659668, "grad_norm": 0.49557754, "learning_rate": 7.05e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87902048, "epoch": 0.38504408, "global_step/max_steps": "950/2468", "percentage": "38.49%", "elapsed_time": "3h 49m 47s", "remaining_time": "6h 7m 10s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068904} +{"loss": 0.3338623, "grad_norm": 0.41627458, "learning_rate": 7.02e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8885435, "epoch": 0.38707063, "global_step/max_steps": "955/2468", "percentage": "38.70%", "elapsed_time": "3h 50m 58s", "remaining_time": "6h 5m 55s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068912} +{"loss": 0.33891602, "grad_norm": 0.36545023, "learning_rate": 6.99e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88697806, "epoch": 0.38909717, "global_step/max_steps": "960/2468", "percentage": "38.90%", "elapsed_time": "3h 52m 6s", "remaining_time": "6h 4m 35s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068935} +{"loss": 0.3736084, "grad_norm": 0.39636111, "learning_rate": 6.96e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87557117, "epoch": 0.39112372, "global_step/max_steps": "965/2468", "percentage": "39.10%", "elapsed_time": "3h 53m 14s", "remaining_time": "6h 3m 16s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068957} +{"loss": 0.3814209, "grad_norm": 0.51710379, "learning_rate": 6.93e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87426171, "epoch": 0.39315027, "global_step/max_steps": "970/2468", "percentage": "39.30%", "elapsed_time": "3h 54m 19s", "remaining_time": "6h 1m 52s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068994} +{"loss": 0.3652832, "grad_norm": 0.54047084, "learning_rate": 6.9e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88294119, "epoch": 0.39517682, "global_step/max_steps": "975/2468", "percentage": "39.51%", "elapsed_time": "3h 55m 27s", "remaining_time": "6h 0m 33s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069013} +{"loss": 0.39111328, "grad_norm": 0.52768636, "learning_rate": 6.87e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.870732, "epoch": 0.39720336, "global_step/max_steps": "980/2468", "percentage": "39.71%", "elapsed_time": "3h 56m 31s", "remaining_time": "5h 59m 8s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069053} +{"loss": 0.38103027, "grad_norm": 0.51233608, "learning_rate": 6.84e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87388389, "epoch": 0.39922991, "global_step/max_steps": "985/2468", "percentage": "39.91%", "elapsed_time": "3h 57m 39s", "remaining_time": "5h 57m 48s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069076} +{"loss": 0.33869629, "grad_norm": 0.37030548, "learning_rate": 6.81e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88851877, "epoch": 0.40125646, "global_step/max_steps": "990/2468", "percentage": "40.11%", "elapsed_time": "3h 58m 49s", "remaining_time": "5h 56m 33s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069087} +{"loss": 0.36213379, "grad_norm": 0.51298159, "learning_rate": 6.78e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87859358, "epoch": 0.40328301, "global_step/max_steps": "995/2468", "percentage": "40.32%", "elapsed_time": "3h 59m 56s", "remaining_time": "5h 55m 12s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069114} +{"loss": 0.36828613, "grad_norm": 0.48550746, "learning_rate": 6.74e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87822306, "epoch": 0.40530956, "global_step/max_steps": "1000/2468", "percentage": "40.52%", "elapsed_time": "4h 1m 4s", "remaining_time": "5h 53m 54s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069133} +{"loss": 0.33730469, "grad_norm": 0.36008275, "learning_rate": 6.71e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88562873, "epoch": 0.4073361, "global_step/max_steps": "1005/2468", "percentage": "40.72%", "elapsed_time": "4h 2m 15s", "remaining_time": "5h 52m 39s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069141} +{"loss": 0.35632324, "grad_norm": 0.37603188, "learning_rate": 6.68e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8803879, "epoch": 0.40936265, "global_step/max_steps": "1010/2468", "percentage": "40.92%", "elapsed_time": "4h 3m 23s", "remaining_time": "5h 51m 20s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069163} +{"loss": 0.34973145, "grad_norm": 0.37092805, "learning_rate": 6.65e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88699763, "epoch": 0.4113892, "global_step/max_steps": "1015/2468", "percentage": "41.13%", "elapsed_time": "4h 4m 34s", "remaining_time": "5h 50m 6s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069168} +{"loss": 0.38139648, "grad_norm": 0.40523756, "learning_rate": 6.62e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87300418, "epoch": 0.41341575, "global_step/max_steps": "1020/2468", "percentage": "41.33%", "elapsed_time": "4h 5m 41s", "remaining_time": "5h 48m 46s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069194} +{"loss": 0.35158691, "grad_norm": 0.39146504, "learning_rate": 6.59e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88377717, "epoch": 0.41544229, "global_step/max_steps": "1025/2468", "percentage": "41.53%", "elapsed_time": "4h 6m 49s", "remaining_time": "5h 47m 29s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069211} +{"loss": 0.34807129, "grad_norm": 0.40089619, "learning_rate": 6.56e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88394271, "epoch": 0.41746884, "global_step/max_steps": "1030/2468", "percentage": "41.73%", "elapsed_time": "4h 7m 58s", "remaining_time": "5h 46m 11s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069229} +{"loss": 0.35285645, "grad_norm": 0.47946051, "learning_rate": 6.53e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88430521, "epoch": 0.41949539, "global_step/max_steps": "1035/2468", "percentage": "41.94%", "elapsed_time": "4h 9m 9s", "remaining_time": "5h 44m 58s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069232} +{"loss": 0.33486328, "grad_norm": 0.53629076, "learning_rate": 6.5e-06, "tIoU": 0.99995422, "tRecall_30": 0.99995365, "tRecall_50": 0.99995365, "tRecall_70": 0.99995365, "token_acc": 0.89226357, "epoch": 0.42152194, "global_step/max_steps": "1040/2468", "percentage": "42.14%", "elapsed_time": "4h 10m 19s", "remaining_time": "5h 43m 42s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069245} +{"loss": 0.37939453, "grad_norm": 0.3557111, "learning_rate": 6.46e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8778559, "epoch": 0.42354849, "global_step/max_steps": "1045/2468", "percentage": "42.34%", "elapsed_time": "4h 11m 27s", "remaining_time": "5h 42m 25s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069262} +{"loss": 0.34943848, "grad_norm": 0.36828193, "learning_rate": 6.43e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88325119, "epoch": 0.42557503, "global_step/max_steps": "1050/2468", "percentage": "42.54%", "elapsed_time": "4h 12m 40s", "remaining_time": "5h 41m 13s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069261} +{"loss": 0.32160645, "grad_norm": 0.35322103, "learning_rate": 6.4e-06, "tIoU": 0.99375, "tRecall_30": 0.99375, "tRecall_50": 0.99375, "tRecall_70": 0.99375, "token_acc": 0.89660783, "epoch": 0.42760158, "global_step/max_steps": "1055/2468", "percentage": "42.75%", "elapsed_time": "4h 13m 54s", "remaining_time": "5h 40m 3s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069252} +{"loss": 0.35319824, "grad_norm": 0.50316286, "learning_rate": 6.37e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88161625, "epoch": 0.42962813, "global_step/max_steps": "1060/2468", "percentage": "42.95%", "elapsed_time": "4h 15m 4s", "remaining_time": "5h 38m 49s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069259} +{"loss": 0.33925781, "grad_norm": 0.40962631, "learning_rate": 6.34e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88646888, "epoch": 0.43165468, "global_step/max_steps": "1065/2468", "percentage": "43.15%", "elapsed_time": "4h 16m 14s", "remaining_time": "5h 37m 33s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069271} +{"loss": 0.35974121, "grad_norm": 0.41023844, "learning_rate": 6.31e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87997943, "epoch": 0.43368122, "global_step/max_steps": "1070/2468", "percentage": "43.35%", "elapsed_time": "4h 17m 23s", "remaining_time": "5h 36m 18s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069283} +{"loss": 0.36403809, "grad_norm": 0.46520615, "learning_rate": 6.28e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8776117, "epoch": 0.43570777, "global_step/max_steps": "1075/2468", "percentage": "43.56%", "elapsed_time": "4h 18m 31s", "remaining_time": "5h 35m 0s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069303} +{"loss": 0.36923828, "grad_norm": 0.48229355, "learning_rate": 6.24e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87743881, "epoch": 0.43773432, "global_step/max_steps": "1080/2468", "percentage": "43.76%", "elapsed_time": "4h 19m 41s", "remaining_time": "5h 33m 44s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069314} +{"loss": 0.36081543, "grad_norm": 0.48839682, "learning_rate": 6.21e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87909623, "epoch": 0.43976087, "global_step/max_steps": "1085/2468", "percentage": "43.96%", "elapsed_time": "4h 20m 50s", "remaining_time": "5h 32m 28s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069327} +{"loss": 0.35488281, "grad_norm": 0.43007761, "learning_rate": 6.18e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88378817, "epoch": 0.44178742, "global_step/max_steps": "1090/2468", "percentage": "44.17%", "elapsed_time": "4h 22m 0s", "remaining_time": "5h 31m 14s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069336} +{"loss": 0.34707031, "grad_norm": 0.48889282, "learning_rate": 6.15e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88543773, "epoch": 0.44381396, "global_step/max_steps": "1095/2468", "percentage": "44.37%", "elapsed_time": "4h 23m 9s", "remaining_time": "5h 29m 58s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.06935} +{"loss": 0.39643555, "grad_norm": 0.46663091, "learning_rate": 6.12e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86790604, "epoch": 0.44584051, "global_step/max_steps": "1100/2468", "percentage": "44.57%", "elapsed_time": "4h 24m 18s", "remaining_time": "5h 28m 41s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069366} +{"loss": 0.35366211, "grad_norm": 0.42913663, "learning_rate": 6.08e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88267471, "epoch": 0.44786706, "global_step/max_steps": "1105/2468", "percentage": "44.77%", "elapsed_time": "4h 25m 28s", "remaining_time": "5h 27m 27s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069372} +{"loss": 0.33896484, "grad_norm": 0.48241073, "learning_rate": 6.05e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88578389, "epoch": 0.44989361, "global_step/max_steps": "1110/2468", "percentage": "44.98%", "elapsed_time": "4h 26m 37s", "remaining_time": "5h 26m 12s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069384} +{"loss": 0.35778809, "grad_norm": 0.42767546, "learning_rate": 6.02e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8800625, "epoch": 0.45192015, "global_step/max_steps": "1115/2468", "percentage": "45.18%", "elapsed_time": "4h 27m 46s", "remaining_time": "5h 24m 56s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069398} +{"loss": 0.32001953, "grad_norm": 0.48337513, "learning_rate": 5.99e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89403945, "epoch": 0.4539467, "global_step/max_steps": "1120/2468", "percentage": "45.38%", "elapsed_time": "4h 28m 58s", "remaining_time": "5h 23m 43s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.0694} +{"loss": 0.35256348, "grad_norm": 0.42286775, "learning_rate": 5.96e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88307795, "epoch": 0.45597325, "global_step/max_steps": "1125/2468", "percentage": "45.58%", "elapsed_time": "4h 30m 7s", "remaining_time": "5h 22m 28s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069412} +{"loss": 0.36379395, "grad_norm": 0.42765027, "learning_rate": 5.92e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8791735, "epoch": 0.4579998, "global_step/max_steps": "1130/2468", "percentage": "45.79%", "elapsed_time": "4h 31m 15s", "remaining_time": "5h 21m 11s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069429} +{"loss": 0.35959473, "grad_norm": 0.46918392, "learning_rate": 5.89e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88081752, "epoch": 0.46002635, "global_step/max_steps": "1135/2468", "percentage": "45.99%", "elapsed_time": "4h 32m 25s", "remaining_time": "5h 19m 56s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069439} +{"loss": 0.3795166, "grad_norm": 0.43709153, "learning_rate": 5.86e-06, "tIoU": 0.99375, "tRecall_30": 0.99375, "tRecall_50": 0.99375, "tRecall_70": 0.99375, "token_acc": 0.87292802, "epoch": 0.46205289, "global_step/max_steps": "1140/2468", "percentage": "46.19%", "elapsed_time": "4h 33m 32s", "remaining_time": "5h 18m 39s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069459} +{"loss": 0.38730469, "grad_norm": 0.42061505, "learning_rate": 5.83e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87123199, "epoch": 0.46407944, "global_step/max_steps": "1145/2468", "percentage": "46.39%", "elapsed_time": "4h 34m 41s", "remaining_time": "5h 17m 23s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069471} +{"loss": 0.33710937, "grad_norm": 0.31676665, "learning_rate": 5.79e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88657541, "epoch": 0.46610599, "global_step/max_steps": "1150/2468", "percentage": "46.60%", "elapsed_time": "4h 35m 52s", "remaining_time": "5h 16m 11s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069474} +{"loss": 0.36435547, "grad_norm": 0.45327589, "learning_rate": 5.76e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87796405, "epoch": 0.46813254, "global_step/max_steps": "1155/2468", "percentage": "46.80%", "elapsed_time": "4h 37m 1s", "remaining_time": "5h 14m 55s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069488} +{"loss": 0.38969727, "grad_norm": 0.43134612, "learning_rate": 5.73e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87004036, "epoch": 0.47015908, "global_step/max_steps": "1160/2468", "percentage": "47.00%", "elapsed_time": "4h 38m 10s", "remaining_time": "5h 13m 39s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069501} +{"loss": 0.37653809, "grad_norm": 0.4814443, "learning_rate": 5.7e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87576198, "epoch": 0.47218563, "global_step/max_steps": "1165/2468", "percentage": "47.20%", "elapsed_time": "4h 39m 18s", "remaining_time": "5h 12m 23s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069516} +{"loss": 0.33996582, "grad_norm": 0.35136965, "learning_rate": 5.66e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88930984, "epoch": 0.47421218, "global_step/max_steps": "1170/2468", "percentage": "47.41%", "elapsed_time": "4h 40m 28s", "remaining_time": "5h 11m 9s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069526} +{"loss": 0.35109863, "grad_norm": 0.40614352, "learning_rate": 5.63e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88339635, "epoch": 0.47623873, "global_step/max_steps": "1175/2468", "percentage": "47.61%", "elapsed_time": "4h 41m 46s", "remaining_time": "5h 10m 4s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069498} +{"loss": 0.36462402, "grad_norm": 0.55005044, "learning_rate": 5.6e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87879464, "epoch": 0.47826528, "global_step/max_steps": "1180/2468", "percentage": "47.81%", "elapsed_time": "4h 42m 56s", "remaining_time": "5h 8m 49s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069509} +{"loss": 0.35324707, "grad_norm": 0.5975911, "learning_rate": 5.57e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88575544, "epoch": 0.48029182, "global_step/max_steps": "1185/2468", "percentage": "48.01%", "elapsed_time": "4h 44m 6s", "remaining_time": "5h 7m 36s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069515} +{"loss": 0.32177734, "grad_norm": 0.34431791, "learning_rate": 5.53e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89159466, "epoch": 0.48231837, "global_step/max_steps": "1190/2468", "percentage": "48.22%", "elapsed_time": "4h 45m 16s", "remaining_time": "5h 6m 22s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069523} +{"loss": 0.34663086, "grad_norm": 0.33990136, "learning_rate": 5.5e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88623325, "epoch": 0.48434492, "global_step/max_steps": "1195/2468", "percentage": "48.42%", "elapsed_time": "4h 46m 25s", "remaining_time": "5h 5m 7s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069534} +{"loss": 0.36721191, "grad_norm": 0.49846762, "learning_rate": 5.47e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87703629, "epoch": 0.48637147, "global_step/max_steps": "1200/2468", "percentage": "48.62%", "elapsed_time": "4h 47m 33s", "remaining_time": "5h 3m 51s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.06955} +{"loss": 0.36760254, "grad_norm": 0.36075729, "learning_rate": 5.44e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87862407, "epoch": 0.48839801, "global_step/max_steps": "1205/2468", "percentage": "48.82%", "elapsed_time": "4h 48m 41s", "remaining_time": "5h 2m 34s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069568} +{"loss": 0.37912598, "grad_norm": 0.40261912, "learning_rate": 5.4e-06, "tIoU": 0.99974995, "tRecall_30": 0.99955368, "tRecall_50": 0.99955368, "tRecall_70": 0.99955368, "token_acc": 0.87413752, "epoch": 0.49042456, "global_step/max_steps": "1210/2468", "percentage": "49.03%", "elapsed_time": "4h 49m 47s", "remaining_time": "5h 1m 17s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.06959} +{"loss": 0.33327637, "grad_norm": 0.41584486, "learning_rate": 5.37e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89254858, "epoch": 0.49245111, "global_step/max_steps": "1215/2468", "percentage": "49.23%", "elapsed_time": "4h 50m 57s", "remaining_time": "5h 0m 3s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069599} +{"loss": 0.34587402, "grad_norm": 0.40428379, "learning_rate": 5.34e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88489292, "epoch": 0.49447766, "global_step/max_steps": "1220/2468", "percentage": "49.43%", "elapsed_time": "4h 52m 6s", "remaining_time": "4h 58m 48s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069609} +{"loss": 0.33461914, "grad_norm": 0.38991839, "learning_rate": 5.31e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88714135, "epoch": 0.49650421, "global_step/max_steps": "1225/2468", "percentage": "49.64%", "elapsed_time": "4h 53m 15s", "remaining_time": "4h 57m 34s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069619} +{"loss": 0.38364258, "grad_norm": 0.38406676, "learning_rate": 5.27e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87123142, "epoch": 0.49853075, "global_step/max_steps": "1230/2468", "percentage": "49.84%", "elapsed_time": "4h 54m 21s", "remaining_time": "4h 56m 16s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069643} +{"loss": 0.38122559, "grad_norm": 0.47551617, "learning_rate": 5.24e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87354391, "epoch": 0.5005573, "global_step/max_steps": "1235/2468", "percentage": "50.04%", "elapsed_time": "4h 55m 28s", "remaining_time": "4h 54m 59s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069662} +{"loss": 0.3880127, "grad_norm": 0.48450133, "learning_rate": 5.21e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87062815, "epoch": 0.50258385, "global_step/max_steps": "1240/2468", "percentage": "50.24%", "elapsed_time": "4h 56m 34s", "remaining_time": "4h 53m 41s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069686} +{"loss": 0.3583252, "grad_norm": 0.44317359, "learning_rate": 5.17e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88041287, "epoch": 0.5046104, "global_step/max_steps": "1245/2468", "percentage": "50.45%", "elapsed_time": "4h 57m 40s", "remaining_time": "4h 52m 24s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069708} +{"loss": 0.35356445, "grad_norm": 0.4925409, "learning_rate": 5.14e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88157408, "epoch": 0.50663694, "global_step/max_steps": "1250/2468", "percentage": "50.65%", "elapsed_time": "4h 58m 46s", "remaining_time": "4h 51m 7s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069729} +{"loss": 0.36533203, "grad_norm": 0.39856628, "learning_rate": 5.11e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87759122, "epoch": 0.50866349, "global_step/max_steps": "1255/2468", "percentage": "50.85%", "elapsed_time": "4h 59m 53s", "remaining_time": "4h 49m 51s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069748} +{"loss": 0.35595703, "grad_norm": 0.41317877, "learning_rate": 5.08e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88166952, "epoch": 0.51069004, "global_step/max_steps": "1260/2468", "percentage": "51.05%", "elapsed_time": "5h 1m 0s", "remaining_time": "4h 48m 35s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069766} +{"loss": 0.35827637, "grad_norm": 0.4024196, "learning_rate": 5.04e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88037931, "epoch": 0.51271659, "global_step/max_steps": "1265/2468", "percentage": "51.26%", "elapsed_time": "5h 2m 7s", "remaining_time": "4h 47m 18s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069784} +{"loss": 0.36027832, "grad_norm": 0.51328337, "learning_rate": 5.01e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87983864, "epoch": 0.51474314, "global_step/max_steps": "1270/2468", "percentage": "51.46%", "elapsed_time": "5h 3m 13s", "remaining_time": "4h 46m 2s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069805} +{"loss": 0.37612305, "grad_norm": 0.44748247, "learning_rate": 4.98e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87580466, "epoch": 0.51676968, "global_step/max_steps": "1275/2468", "percentage": "51.66%", "elapsed_time": "5h 4m 21s", "remaining_time": "4h 44m 46s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069821} +{"loss": 0.35078125, "grad_norm": 0.39678398, "learning_rate": 4.94e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88348376, "epoch": 0.51879623, "global_step/max_steps": "1280/2468", "percentage": "51.86%", "elapsed_time": "5h 5m 29s", "remaining_time": "4h 43m 32s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069833} +{"loss": 0.37058105, "grad_norm": 0.3833645, "learning_rate": 4.91e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87534509, "epoch": 0.52082278, "global_step/max_steps": "1285/2468", "percentage": "52.07%", "elapsed_time": "5h 6m 37s", "remaining_time": "4h 42m 17s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069845} +{"loss": 0.33742676, "grad_norm": 0.343647, "learning_rate": 4.88e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88646204, "epoch": 0.52284933, "global_step/max_steps": "1290/2468", "percentage": "52.27%", "elapsed_time": "5h 7m 48s", "remaining_time": "4h 41m 5s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069848} +{"loss": 0.32480469, "grad_norm": 0.65442449, "learning_rate": 4.85e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89128192, "epoch": 0.52487587, "global_step/max_steps": "1295/2468", "percentage": "52.47%", "elapsed_time": "5h 8m 58s", "remaining_time": "4h 39m 52s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069855} +{"loss": 0.34553223, "grad_norm": 0.37827507, "learning_rate": 4.81e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.884704, "epoch": 0.52690242, "global_step/max_steps": "1300/2468", "percentage": "52.67%", "elapsed_time": "5h 10m 7s", "remaining_time": "4h 38m 38s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069864} +{"loss": 0.33308105, "grad_norm": 0.40097532, "learning_rate": 4.78e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8884987, "epoch": 0.52892897, "global_step/max_steps": "1305/2468", "percentage": "52.88%", "elapsed_time": "5h 11m 17s", "remaining_time": "4h 37m 25s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069869} +{"loss": 0.3642334, "grad_norm": 0.46430066, "learning_rate": 4.75e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87839659, "epoch": 0.53095552, "global_step/max_steps": "1310/2468", "percentage": "53.08%", "elapsed_time": "5h 12m 23s", "remaining_time": "4h 36m 8s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069891} +{"loss": 0.36647949, "grad_norm": 0.49786872, "learning_rate": 4.71e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87757985, "epoch": 0.53298207, "global_step/max_steps": "1315/2468", "percentage": "53.28%", "elapsed_time": "5h 13m 51s", "remaining_time": "4h 35m 11s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.06983} +{"loss": 0.33859863, "grad_norm": 0.46610528, "learning_rate": 4.68e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88565989, "epoch": 0.53500861, "global_step/max_steps": "1320/2468", "percentage": "53.48%", "elapsed_time": "5h 15m 0s", "remaining_time": "4h 33m 57s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.06984} +{"loss": 0.38586426, "grad_norm": 0.45692897, "learning_rate": 4.65e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87247689, "epoch": 0.53703516, "global_step/max_steps": "1325/2468", "percentage": "53.69%", "elapsed_time": "5h 16m 6s", "remaining_time": "4h 32m 41s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.06986} +{"loss": 0.37216797, "grad_norm": 0.53581679, "learning_rate": 4.62e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87785961, "epoch": 0.53906171, "global_step/max_steps": "1330/2468", "percentage": "53.89%", "elapsed_time": "5h 17m 14s", "remaining_time": "4h 31m 26s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069874} +{"loss": 0.3567627, "grad_norm": 0.46108669, "learning_rate": 4.58e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88218973, "epoch": 0.54108826, "global_step/max_steps": "1335/2468", "percentage": "54.09%", "elapsed_time": "5h 18m 22s", "remaining_time": "4h 30m 12s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069885} +{"loss": 0.37089844, "grad_norm": 0.47000489, "learning_rate": 4.55e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87475639, "epoch": 0.5431148, "global_step/max_steps": "1340/2468", "percentage": "54.29%", "elapsed_time": "5h 19m 29s", "remaining_time": "4h 28m 56s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069904} +{"loss": 0.37182617, "grad_norm": 0.44782415, "learning_rate": 4.52e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87731851, "epoch": 0.54514135, "global_step/max_steps": "1345/2468", "percentage": "54.50%", "elapsed_time": "5h 20m 35s", "remaining_time": "4h 27m 40s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069922} +{"loss": 0.35019531, "grad_norm": 0.40902999, "learning_rate": 4.49e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88314892, "epoch": 0.5471679, "global_step/max_steps": "1350/2468", "percentage": "54.70%", "elapsed_time": "5h 21m 45s", "remaining_time": "4h 26m 27s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069928} +{"loss": 0.34912109, "grad_norm": 0.46472162, "learning_rate": 4.45e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88302124, "epoch": 0.54919445, "global_step/max_steps": "1355/2468", "percentage": "54.90%", "elapsed_time": "5h 22m 53s", "remaining_time": "4h 25m 13s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.06994} +{"loss": 0.35053711, "grad_norm": 0.54915273, "learning_rate": 4.42e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88145083, "epoch": 0.551221, "global_step/max_steps": "1360/2468", "percentage": "55.11%", "elapsed_time": "5h 24m 2s", "remaining_time": "4h 24m 0s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069949} +{"loss": 0.35644531, "grad_norm": 0.39852718, "learning_rate": 4.39e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88100571, "epoch": 0.55324754, "global_step/max_steps": "1365/2468", "percentage": "55.31%", "elapsed_time": "5h 25m 9s", "remaining_time": "4h 22m 45s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069965} +{"loss": 0.3387207, "grad_norm": 0.38414013, "learning_rate": 4.36e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88545921, "epoch": 0.55527409, "global_step/max_steps": "1370/2468", "percentage": "55.51%", "elapsed_time": "5h 26m 16s", "remaining_time": "4h 21m 30s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.06998} +{"loss": 0.37338867, "grad_norm": 0.46751761, "learning_rate": 4.32e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87454562, "epoch": 0.55730064, "global_step/max_steps": "1375/2468", "percentage": "55.71%", "elapsed_time": "5h 27m 22s", "remaining_time": "4h 20m 13s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070002} +{"loss": 0.38967285, "grad_norm": 0.41305846, "learning_rate": 4.29e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.86900928, "epoch": 0.55932719, "global_step/max_steps": "1380/2468", "percentage": "55.92%", "elapsed_time": "5h 28m 27s", "remaining_time": "4h 18m 57s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070024} +{"loss": 0.34763184, "grad_norm": 0.45166895, "learning_rate": 4.26e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88387763, "epoch": 0.56135373, "global_step/max_steps": "1385/2468", "percentage": "56.12%", "elapsed_time": "5h 29m 34s", "remaining_time": "4h 17m 42s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070039} +{"loss": 0.36848145, "grad_norm": 0.3921133, "learning_rate": 4.23e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87690611, "epoch": 0.56338028, "global_step/max_steps": "1390/2468", "percentage": "56.32%", "elapsed_time": "5h 30m 40s", "remaining_time": "4h 16m 27s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070058} +{"loss": 0.35566406, "grad_norm": 0.68565559, "learning_rate": 4.19e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87954529, "epoch": 0.56540683, "global_step/max_steps": "1395/2468", "percentage": "56.52%", "elapsed_time": "5h 31m 47s", "remaining_time": "4h 15m 12s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070073} +{"loss": 0.3482666, "grad_norm": 0.44158694, "learning_rate": 4.16e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88454783, "epoch": 0.56743338, "global_step/max_steps": "1400/2468", "percentage": "56.73%", "elapsed_time": "5h 32m 56s", "remaining_time": "4h 13m 59s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070081} +{"loss": 0.3590332, "grad_norm": 0.44592997, "learning_rate": 4.13e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87891149, "epoch": 0.56945993, "global_step/max_steps": "1405/2468", "percentage": "56.93%", "elapsed_time": "5h 34m 6s", "remaining_time": "4h 12m 46s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070089} +{"loss": 0.34919434, "grad_norm": 0.42004541, "learning_rate": 4.1e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88224072, "epoch": 0.57148647, "global_step/max_steps": "1410/2468", "percentage": "57.13%", "elapsed_time": "5h 35m 15s", "remaining_time": "4h 11m 33s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070095} +{"loss": 0.36396484, "grad_norm": 0.34816152, "learning_rate": 4.06e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8796527, "epoch": 0.57351302, "global_step/max_steps": "1415/2468", "percentage": "57.33%", "elapsed_time": "5h 36m 24s", "remaining_time": "4h 10m 20s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070103} +{"loss": 0.36621094, "grad_norm": 0.57756841, "learning_rate": 4.03e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87836377, "epoch": 0.57553957, "global_step/max_steps": "1420/2468", "percentage": "57.54%", "elapsed_time": "5h 37m 32s", "remaining_time": "4h 9m 6s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070115} +{"loss": 0.36938477, "grad_norm": 0.39185187, "learning_rate": 4e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87746935, "epoch": 0.57756612, "global_step/max_steps": "1425/2468", "percentage": "57.74%", "elapsed_time": "5h 38m 40s", "remaining_time": "4h 7m 53s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070127} +{"loss": 0.34960938, "grad_norm": 0.49676889, "learning_rate": 3.97e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88165922, "epoch": 0.57959266, "global_step/max_steps": "1430/2468", "percentage": "57.94%", "elapsed_time": "5h 39m 48s", "remaining_time": "4h 6m 39s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070136} +{"loss": 0.36030273, "grad_norm": 0.42753616, "learning_rate": 3.93e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88036087, "epoch": 0.58161921, "global_step/max_steps": "1435/2468", "percentage": "58.14%", "elapsed_time": "5h 41m 0s", "remaining_time": "4h 5m 28s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070135} +{"loss": 0.35256348, "grad_norm": 0.4185347, "learning_rate": 3.9e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88114389, "epoch": 0.58364576, "global_step/max_steps": "1440/2468", "percentage": "58.35%", "elapsed_time": "5h 42m 7s", "remaining_time": "4h 4m 14s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07015} +{"loss": 0.37126465, "grad_norm": 0.38084778, "learning_rate": 3.87e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87503128, "epoch": 0.58567231, "global_step/max_steps": "1445/2468", "percentage": "58.55%", "elapsed_time": "5h 43m 15s", "remaining_time": "4h 3m 0s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07016} +{"loss": 0.35402832, "grad_norm": 0.3081311, "learning_rate": 3.84e-06, "tIoU": 0.99977913, "tRecall_30": 0.99970236, "tRecall_50": 0.99970236, "tRecall_70": 0.99970236, "token_acc": 0.88334775, "epoch": 0.58769886, "global_step/max_steps": "1450/2468", "percentage": "58.75%", "elapsed_time": "5h 44m 23s", "remaining_time": "4h 1m 47s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070171} +{"loss": 0.35476074, "grad_norm": 0.37281391, "learning_rate": 3.81e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87996705, "epoch": 0.5897254, "global_step/max_steps": "1455/2468", "percentage": "58.95%", "elapsed_time": "5h 45m 33s", "remaining_time": "4h 0m 35s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070176} +{"loss": 0.35473633, "grad_norm": 0.54544109, "learning_rate": 3.78e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88180268, "epoch": 0.59175195, "global_step/max_steps": "1460/2468", "percentage": "59.16%", "elapsed_time": "5h 46m 40s", "remaining_time": "3h 59m 20s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070192} +{"loss": 0.34846191, "grad_norm": 0.33592007, "learning_rate": 3.74e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88168693, "epoch": 0.5937785, "global_step/max_steps": "1465/2468", "percentage": "59.36%", "elapsed_time": "5h 47m 49s", "remaining_time": "3h 58m 7s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070199} +{"loss": 0.36337891, "grad_norm": 0.48660961, "learning_rate": 3.71e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.876798, "epoch": 0.59580505, "global_step/max_steps": "1470/2468", "percentage": "59.56%", "elapsed_time": "5h 48m 57s", "remaining_time": "3h 56m 54s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07021} +{"loss": 0.35253906, "grad_norm": 0.42000899, "learning_rate": 3.68e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8828759, "epoch": 0.59783159, "global_step/max_steps": "1475/2468", "percentage": "59.76%", "elapsed_time": "5h 50m 6s", "remaining_time": "3h 55m 41s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070216} +{"loss": 0.31704102, "grad_norm": 0.44496635, "learning_rate": 3.65e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89417195, "epoch": 0.59985814, "global_step/max_steps": "1480/2468", "percentage": "59.97%", "elapsed_time": "5h 51m 16s", "remaining_time": "3h 54m 30s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07022} +{"loss": 0.36481934, "grad_norm": 0.44330952, "learning_rate": 3.62e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87796253, "epoch": 0.60188469, "global_step/max_steps": "1485/2468", "percentage": "60.17%", "elapsed_time": "5h 52m 23s", "remaining_time": "3h 53m 15s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070235} +{"loss": 0.33647461, "grad_norm": 0.31316468, "learning_rate": 3.59e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88740356, "epoch": 0.60391124, "global_step/max_steps": "1490/2468", "percentage": "60.37%", "elapsed_time": "5h 53m 30s", "remaining_time": "3h 52m 2s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070248} +{"loss": 0.34521484, "grad_norm": 0.34361219, "learning_rate": 3.55e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.883073, "epoch": 0.60593778, "global_step/max_steps": "1495/2468", "percentage": "60.58%", "elapsed_time": "5h 54m 37s", "remaining_time": "3h 50m 48s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070261} +{"loss": 0.35839844, "grad_norm": 0.3756561, "learning_rate": 3.52e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88217011, "epoch": 0.60796433, "global_step/max_steps": "1500/2468", "percentage": "60.78%", "elapsed_time": "5h 55m 47s", "remaining_time": "3h 49m 36s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070266} +{"loss": 0.36647949, "grad_norm": 0.4399938, "learning_rate": 3.49e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87742885, "epoch": 0.60999088, "global_step/max_steps": "1505/2468", "percentage": "60.98%", "elapsed_time": "5h 56m 54s", "remaining_time": "3h 48m 22s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070279} +{"loss": 0.34836426, "grad_norm": 0.48006973, "learning_rate": 3.46e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8825046, "epoch": 0.61201743, "global_step/max_steps": "1510/2468", "percentage": "61.18%", "elapsed_time": "5h 58m 3s", "remaining_time": "3h 47m 9s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070286} +{"loss": 0.33679199, "grad_norm": 0.37062064, "learning_rate": 3.43e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88699972, "epoch": 0.61404398, "global_step/max_steps": "1515/2468", "percentage": "61.39%", "elapsed_time": "5h 59m 13s", "remaining_time": "3h 45m 57s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070291} +{"loss": 0.33410645, "grad_norm": 0.45247993, "learning_rate": 3.4e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88875789, "epoch": 0.61607052, "global_step/max_steps": "1520/2468", "percentage": "61.59%", "elapsed_time": "6h 0m 22s", "remaining_time": "3h 44m 45s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070298} +{"loss": 0.35632324, "grad_norm": 0.37689427, "learning_rate": 3.37e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88232012, "epoch": 0.61809707, "global_step/max_steps": "1525/2468", "percentage": "61.79%", "elapsed_time": "6h 1m 30s", "remaining_time": "3h 43m 32s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070309} +{"loss": 0.35014648, "grad_norm": 0.40862408, "learning_rate": 3.34e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88242313, "epoch": 0.62012362, "global_step/max_steps": "1530/2468", "percentage": "61.99%", "elapsed_time": "6h 2m 38s", "remaining_time": "3h 42m 19s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070316} +{"loss": 0.34528809, "grad_norm": 0.39199698, "learning_rate": 3.3e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88392774, "epoch": 0.62215017, "global_step/max_steps": "1535/2468", "percentage": "62.20%", "elapsed_time": "6h 3m 48s", "remaining_time": "3h 41m 7s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070322} +{"loss": 0.34035645, "grad_norm": 0.44045988, "learning_rate": 3.27e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88588258, "epoch": 0.62417671, "global_step/max_steps": "1540/2468", "percentage": "62.40%", "elapsed_time": "6h 4m 56s", "remaining_time": "3h 39m 54s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07033} +{"loss": 0.33649902, "grad_norm": 0.33820724, "learning_rate": 3.24e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88622426, "epoch": 0.62620326, "global_step/max_steps": "1545/2468", "percentage": "62.60%", "elapsed_time": "6h 6m 7s", "remaining_time": "3h 38m 43s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070333} +{"loss": 0.32089844, "grad_norm": 0.33277419, "learning_rate": 3.21e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89283382, "epoch": 0.62822981, "global_step/max_steps": "1550/2468", "percentage": "62.80%", "elapsed_time": "6h 7m 19s", "remaining_time": "3h 37m 32s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070329} +{"loss": 0.34240723, "grad_norm": 0.5623495, "learning_rate": 3.18e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88540341, "epoch": 0.63025636, "global_step/max_steps": "1555/2468", "percentage": "63.01%", "elapsed_time": "6h 8m 27s", "remaining_time": "3h 36m 19s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070339} +{"loss": 0.32331543, "grad_norm": 0.4070521, "learning_rate": 3.15e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89196931, "epoch": 0.63228291, "global_step/max_steps": "1560/2468", "percentage": "63.21%", "elapsed_time": "6h 9m 38s", "remaining_time": "3h 35m 8s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070339} +{"loss": 0.32705078, "grad_norm": 0.38860211, "learning_rate": 3.12e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89058537, "epoch": 0.63430945, "global_step/max_steps": "1565/2468", "percentage": "63.41%", "elapsed_time": "6h 10m 48s", "remaining_time": "3h 33m 57s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070341} +{"loss": 0.34023437, "grad_norm": 0.4521631, "learning_rate": 3.09e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88515678, "epoch": 0.636336, "global_step/max_steps": "1570/2468", "percentage": "63.61%", "elapsed_time": "6h 11m 57s", "remaining_time": "3h 32m 45s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070347} +{"loss": 0.3456543, "grad_norm": 0.42403921, "learning_rate": 3.06e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88412031, "epoch": 0.63836255, "global_step/max_steps": "1575/2468", "percentage": "63.82%", "elapsed_time": "6h 13m 5s", "remaining_time": "3h 31m 31s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070359} +{"loss": 0.35246582, "grad_norm": 0.43884194, "learning_rate": 3.03e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88212444, "epoch": 0.6403891, "global_step/max_steps": "1580/2468", "percentage": "64.02%", "elapsed_time": "6h 14m 11s", "remaining_time": "3h 30m 18s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070373} +{"loss": 0.35192871, "grad_norm": 0.38225412, "learning_rate": 3e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88129856, "epoch": 0.64241564, "global_step/max_steps": "1585/2468", "percentage": "64.22%", "elapsed_time": "6h 15m 21s", "remaining_time": "3h 29m 6s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070378} +{"loss": 0.34758301, "grad_norm": 0.44400823, "learning_rate": 2.97e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88434087, "epoch": 0.64444219, "global_step/max_steps": "1590/2468", "percentage": "64.42%", "elapsed_time": "6h 16m 30s", "remaining_time": "3h 27m 54s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070382} +{"loss": 0.34055176, "grad_norm": 0.38897669, "learning_rate": 2.94e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88713165, "epoch": 0.64646874, "global_step/max_steps": "1595/2468", "percentage": "64.63%", "elapsed_time": "6h 17m 40s", "remaining_time": "3h 26m 42s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070386} +{"loss": 0.35910645, "grad_norm": 0.460372, "learning_rate": 2.91e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87846505, "epoch": 0.64849529, "global_step/max_steps": "1600/2468", "percentage": "64.83%", "elapsed_time": "6h 18m 50s", "remaining_time": "3h 25m 31s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070389} +{"loss": 0.36091309, "grad_norm": 0.48420817, "learning_rate": 2.88e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87913785, "epoch": 0.65052184, "global_step/max_steps": "1605/2468", "percentage": "65.03%", "elapsed_time": "6h 20m 1s", "remaining_time": "3h 24m 20s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07039} +{"loss": 0.3239502, "grad_norm": 0.3952024, "learning_rate": 2.85e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89045204, "epoch": 0.65254838, "global_step/max_steps": "1610/2468", "percentage": "65.24%", "elapsed_time": "6h 21m 12s", "remaining_time": "3h 23m 9s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07039} +{"loss": 0.34152832, "grad_norm": 0.45618477, "learning_rate": 2.82e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88606181, "epoch": 0.65457493, "global_step/max_steps": "1615/2468", "percentage": "65.44%", "elapsed_time": "6h 22m 22s", "remaining_time": "3h 21m 57s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070394} +{"loss": 0.35822754, "grad_norm": 0.47599977, "learning_rate": 2.79e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87943477, "epoch": 0.65660148, "global_step/max_steps": "1620/2468", "percentage": "65.64%", "elapsed_time": "6h 23m 28s", "remaining_time": "3h 20m 43s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070409} +{"loss": 0.32006836, "grad_norm": 0.47110984, "learning_rate": 2.76e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89251887, "epoch": 0.65862803, "global_step/max_steps": "1625/2468", "percentage": "65.84%", "elapsed_time": "6h 24m 38s", "remaining_time": "3h 19m 32s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070413} +{"loss": 0.32485352, "grad_norm": 0.47064129, "learning_rate": 2.73e-06, "tIoU": 0.99987326, "tRecall_30": 0.99958334, "tRecall_50": 0.99958334, "tRecall_70": 0.99958334, "token_acc": 0.88976957, "epoch": 0.66065457, "global_step/max_steps": "1630/2468", "percentage": "66.05%", "elapsed_time": "6h 25m 47s", "remaining_time": "3h 18m 20s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070418} +{"loss": 0.33527832, "grad_norm": 0.47430021, "learning_rate": 2.7e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88705095, "epoch": 0.66268112, "global_step/max_steps": "1635/2468", "percentage": "66.25%", "elapsed_time": "6h 26m 55s", "remaining_time": "3h 17m 7s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070426} +{"loss": 0.35168457, "grad_norm": 0.42994794, "learning_rate": 2.67e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88157234, "epoch": 0.66470767, "global_step/max_steps": "1640/2468", "percentage": "66.45%", "elapsed_time": "6h 28m 5s", "remaining_time": "3h 15m 56s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070431} +{"loss": 0.31716309, "grad_norm": 0.37945184, "learning_rate": 2.65e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89624192, "epoch": 0.66673422, "global_step/max_steps": "1645/2468", "percentage": "66.65%", "elapsed_time": "6h 29m 16s", "remaining_time": "3h 14m 45s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070429} +{"loss": 0.34768066, "grad_norm": 0.47009155, "learning_rate": 2.62e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8828485, "epoch": 0.66876077, "global_step/max_steps": "1650/2468", "percentage": "66.86%", "elapsed_time": "6h 30m 25s", "remaining_time": "3h 13m 33s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070436} +{"loss": 0.32387695, "grad_norm": 0.32283145, "learning_rate": 2.59e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89169235, "epoch": 0.67078731, "global_step/max_steps": "1655/2468", "percentage": "67.06%", "elapsed_time": "6h 31m 35s", "remaining_time": "3h 12m 22s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070438} +{"loss": 0.33215332, "grad_norm": 0.41415188, "learning_rate": 2.56e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88893928, "epoch": 0.67281386, "global_step/max_steps": "1660/2468", "percentage": "67.26%", "elapsed_time": "6h 32m 44s", "remaining_time": "3h 11m 9s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070446} +{"loss": 0.35144043, "grad_norm": 0.50321579, "learning_rate": 2.53e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88226587, "epoch": 0.67484041, "global_step/max_steps": "1665/2468", "percentage": "67.46%", "elapsed_time": "6h 33m 51s", "remaining_time": "3h 9m 57s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070457} +{"loss": 0.34138184, "grad_norm": 0.54603827, "learning_rate": 2.5e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88648383, "epoch": 0.67686696, "global_step/max_steps": "1670/2468", "percentage": "67.67%", "elapsed_time": "6h 35m 1s", "remaining_time": "3h 8m 45s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070458} +{"loss": 0.32338867, "grad_norm": 0.40518916, "learning_rate": 2.47e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89040252, "epoch": 0.6788935, "global_step/max_steps": "1675/2468", "percentage": "67.87%", "elapsed_time": "6h 36m 11s", "remaining_time": "3h 7m 34s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070463} +{"loss": 0.35354004, "grad_norm": 0.42271355, "learning_rate": 2.45e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88156199, "epoch": 0.68092005, "global_step/max_steps": "1680/2468", "percentage": "68.07%", "elapsed_time": "6h 37m 18s", "remaining_time": "3h 6m 21s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070474} +{"loss": 0.34765625, "grad_norm": 0.48655456, "learning_rate": 2.42e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88285368, "epoch": 0.6829466, "global_step/max_steps": "1685/2468", "percentage": "68.27%", "elapsed_time": "6h 38m 26s", "remaining_time": "3h 5m 9s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070483} +{"loss": 0.35002441, "grad_norm": 0.46620539, "learning_rate": 2.39e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88239264, "epoch": 0.68497315, "global_step/max_steps": "1690/2468", "percentage": "68.48%", "elapsed_time": "6h 39m 34s", "remaining_time": "3h 3m 56s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070492} +{"loss": 0.33989258, "grad_norm": 0.50177115, "learning_rate": 2.36e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88560212, "epoch": 0.6869997, "global_step/max_steps": "1695/2468", "percentage": "68.68%", "elapsed_time": "6h 40m 44s", "remaining_time": "3h 2m 45s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070495} +{"loss": 0.35483398, "grad_norm": 0.4281114, "learning_rate": 2.33e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88172003, "epoch": 0.68902624, "global_step/max_steps": "1700/2468", "percentage": "68.88%", "elapsed_time": "6h 41m 52s", "remaining_time": "3h 1m 33s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070503} +{"loss": 0.34006348, "grad_norm": 0.45597005, "learning_rate": 2.31e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88659034, "epoch": 0.69105279, "global_step/max_steps": "1705/2468", "percentage": "69.08%", "elapsed_time": "6h 42m 59s", "remaining_time": "3h 0m 20s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070514} +{"loss": 0.32783203, "grad_norm": 0.45921934, "learning_rate": 2.28e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88937592, "epoch": 0.69307934, "global_step/max_steps": "1710/2468", "percentage": "69.29%", "elapsed_time": "6h 44m 5s", "remaining_time": "2h 59m 7s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070529} +{"loss": 0.36813965, "grad_norm": 0.50059313, "learning_rate": 2.25e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8778965, "epoch": 0.69510589, "global_step/max_steps": "1715/2468", "percentage": "69.49%", "elapsed_time": "6h 45m 12s", "remaining_time": "2h 57m 54s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070541} +{"loss": 0.29257813, "grad_norm": 0.26108834, "learning_rate": 2.22e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.90554914, "epoch": 0.69713243, "global_step/max_steps": "1720/2468", "percentage": "69.69%", "elapsed_time": "6h 46m 26s", "remaining_time": "2h 56m 45s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070532} +{"loss": 0.35205078, "grad_norm": 0.47036916, "learning_rate": 2.2e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88277837, "epoch": 0.69915898, "global_step/max_steps": "1725/2468", "percentage": "69.89%", "elapsed_time": "6h 47m 32s", "remaining_time": "2h 55m 32s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070544} +{"loss": 0.32626953, "grad_norm": 0.35006201, "learning_rate": 2.17e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88986991, "epoch": 0.70118553, "global_step/max_steps": "1730/2468", "percentage": "70.10%", "elapsed_time": "6h 48m 41s", "remaining_time": "2h 54m 20s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07055} +{"loss": 0.35883789, "grad_norm": 0.39124799, "learning_rate": 2.14e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87903429, "epoch": 0.70321208, "global_step/max_steps": "1735/2468", "percentage": "70.30%", "elapsed_time": "6h 49m 48s", "remaining_time": "2h 53m 8s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070561} +{"loss": 0.35568848, "grad_norm": 0.45080665, "learning_rate": 2.11e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88068469, "epoch": 0.70523863, "global_step/max_steps": "1740/2468", "percentage": "70.50%", "elapsed_time": "6h 50m 55s", "remaining_time": "2h 51m 55s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070572} +{"loss": 0.34519043, "grad_norm": 0.38651094, "learning_rate": 2.09e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88436289, "epoch": 0.70726517, "global_step/max_steps": "1745/2468", "percentage": "70.71%", "elapsed_time": "6h 52m 4s", "remaining_time": "2h 50m 43s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070579} +{"loss": 0.30239258, "grad_norm": 0.46912605, "learning_rate": 2.06e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.90062805, "epoch": 0.70929172, "global_step/max_steps": "1750/2468", "percentage": "70.91%", "elapsed_time": "6h 53m 16s", "remaining_time": "2h 49m 33s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070575} +{"loss": 0.34765625, "grad_norm": 0.46624127, "learning_rate": 2.04e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88271701, "epoch": 0.71131827, "global_step/max_steps": "1755/2468", "percentage": "71.11%", "elapsed_time": "6h 54m 25s", "remaining_time": "2h 48m 21s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07058} +{"loss": 0.34709473, "grad_norm": 0.46779165, "learning_rate": 2.01e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88372727, "epoch": 0.71334482, "global_step/max_steps": "1760/2468", "percentage": "71.31%", "elapsed_time": "6h 55m 33s", "remaining_time": "2h 47m 10s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070587} +{"loss": 0.34707031, "grad_norm": 0.52653086, "learning_rate": 1.98e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88244885, "epoch": 0.71537136, "global_step/max_steps": "1765/2468", "percentage": "71.52%", "elapsed_time": "6h 56m 42s", "remaining_time": "2h 45m 58s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070593} +{"loss": 0.34763184, "grad_norm": 0.58091837, "learning_rate": 1.96e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88548244, "epoch": 0.71739791, "global_step/max_steps": "1770/2468", "percentage": "71.72%", "elapsed_time": "6h 57m 51s", "remaining_time": "2h 44m 47s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070598} +{"loss": 0.34174805, "grad_norm": 0.44178304, "learning_rate": 1.93e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88571819, "epoch": 0.71942446, "global_step/max_steps": "1775/2468", "percentage": "71.92%", "elapsed_time": "6h 58m 58s", "remaining_time": "2h 43m 34s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070609} +{"loss": 0.33847656, "grad_norm": 0.36755958, "learning_rate": 1.9e-06, "tIoU": 0.99898148, "tRecall_30": 0.99963226, "tRecall_50": 0.99963226, "tRecall_70": 0.99963226, "token_acc": 0.88610894, "epoch": 0.72145101, "global_step/max_steps": "1780/2468", "percentage": "72.12%", "elapsed_time": "7h 0m 5s", "remaining_time": "2h 42m 22s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07062} +{"loss": 0.35690918, "grad_norm": 0.47204509, "learning_rate": 1.88e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87930766, "epoch": 0.72347756, "global_step/max_steps": "1785/2468", "percentage": "72.33%", "elapsed_time": "7h 1m 11s", "remaining_time": "2h 41m 9s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070633} +{"loss": 0.3317627, "grad_norm": 0.36413619, "learning_rate": 1.85e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8881009, "epoch": 0.7255041, "global_step/max_steps": "1790/2468", "percentage": "72.53%", "elapsed_time": "7h 2m 20s", "remaining_time": "2h 39m 58s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070639} +{"loss": 0.30683594, "grad_norm": 0.48178297, "learning_rate": 1.83e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89764655, "epoch": 0.72753065, "global_step/max_steps": "1795/2468", "percentage": "72.73%", "elapsed_time": "7h 3m 31s", "remaining_time": "2h 38m 47s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070637} +{"loss": 0.3442627, "grad_norm": 0.44085571, "learning_rate": 1.8e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88330641, "epoch": 0.7295572, "global_step/max_steps": "1800/2468", "percentage": "72.93%", "elapsed_time": "7h 4m 41s", "remaining_time": "2h 37m 36s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07064} +{"loss": 0.3112793, "grad_norm": 0.44399267, "learning_rate": 1.78e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89748594, "epoch": 0.73158375, "global_step/max_steps": "1805/2468", "percentage": "73.14%", "elapsed_time": "7h 5m 52s", "remaining_time": "2h 36m 25s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070638} +{"loss": 0.32043457, "grad_norm": 0.43583795, "learning_rate": 1.75e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8913285, "epoch": 0.73361029, "global_step/max_steps": "1810/2468", "percentage": "73.34%", "elapsed_time": "7h 7m 2s", "remaining_time": "2h 35m 14s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07064} +{"loss": 0.33659668, "grad_norm": 0.36242667, "learning_rate": 1.73e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88815546, "epoch": 0.73563684, "global_step/max_steps": "1815/2468", "percentage": "73.54%", "elapsed_time": "7h 8m 11s", "remaining_time": "2h 34m 3s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070646} +{"loss": 0.34714355, "grad_norm": 0.43046907, "learning_rate": 1.7e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88295728, "epoch": 0.73766339, "global_step/max_steps": "1820/2468", "percentage": "73.74%", "elapsed_time": "7h 9m 19s", "remaining_time": "2h 32m 51s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070654} +{"loss": 0.35683594, "grad_norm": 0.40860116, "learning_rate": 1.68e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88137704, "epoch": 0.73968994, "global_step/max_steps": "1825/2468", "percentage": "73.95%", "elapsed_time": "7h 10m 26s", "remaining_time": "2h 31m 39s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070663} +{"loss": 0.32495117, "grad_norm": 0.57225049, "learning_rate": 1.65e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.889589, "epoch": 0.74171649, "global_step/max_steps": "1830/2468", "percentage": "74.15%", "elapsed_time": "7h 11m 36s", "remaining_time": "2h 30m 28s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070667} +{"loss": 0.34707031, "grad_norm": 0.42056841, "learning_rate": 1.63e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88322198, "epoch": 0.74374303, "global_step/max_steps": "1835/2468", "percentage": "74.35%", "elapsed_time": "7h 12m 44s", "remaining_time": "2h 29m 16s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070672} +{"loss": 0.36586914, "grad_norm": 0.45901892, "learning_rate": 1.61e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87763226, "epoch": 0.74576958, "global_step/max_steps": "1840/2468", "percentage": "74.55%", "elapsed_time": "7h 13m 52s", "remaining_time": "2h 28m 4s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070682} +{"loss": 0.3380127, "grad_norm": 0.47276136, "learning_rate": 1.58e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88798224, "epoch": 0.74779613, "global_step/max_steps": "1845/2468", "percentage": "74.76%", "elapsed_time": "7h 15m 0s", "remaining_time": "2h 26m 53s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070687} +{"loss": 0.3182373, "grad_norm": 0.24706824, "learning_rate": 1.56e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89557098, "epoch": 0.74982268, "global_step/max_steps": "1850/2468", "percentage": "74.96%", "elapsed_time": "7h 16m 13s", "remaining_time": "2h 25m 43s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070682} +{"loss": 0.30827637, "grad_norm": 0.41805166, "learning_rate": 1.53e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89682512, "epoch": 0.75184922, "global_step/max_steps": "1855/2468", "percentage": "75.16%", "elapsed_time": "7h 17m 23s", "remaining_time": "2h 24m 32s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070684} +{"loss": 0.32416992, "grad_norm": 0.38584384, "learning_rate": 1.51e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89085453, "epoch": 0.75387577, "global_step/max_steps": "1860/2468", "percentage": "75.36%", "elapsed_time": "7h 18m 33s", "remaining_time": "2h 23m 21s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070686} +{"loss": 0.34052734, "grad_norm": 0.44953439, "learning_rate": 1.49e-06, "tIoU": 0.99375, "tRecall_30": 0.99375, "tRecall_50": 0.99375, "tRecall_70": 0.99375, "token_acc": 0.88631641, "epoch": 0.75590232, "global_step/max_steps": "1865/2468", "percentage": "75.57%", "elapsed_time": "7h 19m 42s", "remaining_time": "2h 22m 10s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070691} +{"loss": 0.3251709, "grad_norm": 0.40899697, "learning_rate": 1.46e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88942972, "epoch": 0.75792887, "global_step/max_steps": "1870/2468", "percentage": "75.77%", "elapsed_time": "7h 20m 50s", "remaining_time": "2h 20m 58s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070697} +{"loss": 0.34287109, "grad_norm": 0.30098531, "learning_rate": 1.44e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88442334, "epoch": 0.75995542, "global_step/max_steps": "1875/2468", "percentage": "75.97%", "elapsed_time": "7h 21m 59s", "remaining_time": "2h 19m 47s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070702} +{"loss": 0.29780273, "grad_norm": 0.39261687, "learning_rate": 1.42e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.9007301, "epoch": 0.76198196, "global_step/max_steps": "1880/2468", "percentage": "76.18%", "elapsed_time": "7h 23m 10s", "remaining_time": "2h 18m 36s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070701} +{"loss": 0.3418457, "grad_norm": 0.46458375, "learning_rate": 1.39e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88544142, "epoch": 0.76400851, "global_step/max_steps": "1885/2468", "percentage": "76.38%", "elapsed_time": "7h 24m 19s", "remaining_time": "2h 17m 25s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070706} +{"loss": 0.33195801, "grad_norm": 0.43175471, "learning_rate": 1.37e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88707855, "epoch": 0.76603506, "global_step/max_steps": "1890/2468", "percentage": "76.58%", "elapsed_time": "7h 25m 28s", "remaining_time": "2h 16m 14s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07071} +{"loss": 0.33774414, "grad_norm": 0.4843618, "learning_rate": 1.35e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88586995, "epoch": 0.76806161, "global_step/max_steps": "1895/2468", "percentage": "76.78%", "elapsed_time": "7h 26m 37s", "remaining_time": "2h 15m 2s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070715} +{"loss": 0.32731934, "grad_norm": 0.4182682, "learning_rate": 1.33e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89007942, "epoch": 0.77008815, "global_step/max_steps": "1900/2468", "percentage": "76.99%", "elapsed_time": "7h 27m 46s", "remaining_time": "2h 13m 51s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070721} +{"loss": 0.35039063, "grad_norm": 0.52274972, "learning_rate": 1.3e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88224422, "epoch": 0.7721147, "global_step/max_steps": "1905/2468", "percentage": "77.19%", "elapsed_time": "7h 28m 54s", "remaining_time": "2h 12m 40s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070728} +{"loss": 0.32856445, "grad_norm": 0.41229686, "learning_rate": 1.28e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8884929, "epoch": 0.77414125, "global_step/max_steps": "1910/2468", "percentage": "77.39%", "elapsed_time": "7h 30m 3s", "remaining_time": "2h 11m 28s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070732} +{"loss": 0.32509766, "grad_norm": 0.47801197, "learning_rate": 1.26e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89165727, "epoch": 0.7761678, "global_step/max_steps": "1915/2468", "percentage": "77.59%", "elapsed_time": "7h 31m 16s", "remaining_time": "2h 10m 18s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070727} +{"loss": 0.32563477, "grad_norm": 0.41708609, "learning_rate": 1.24e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89099828, "epoch": 0.77819435, "global_step/max_steps": "1920/2468", "percentage": "77.80%", "elapsed_time": "7h 32m 27s", "remaining_time": "2h 9m 8s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070726} +{"loss": 0.34138184, "grad_norm": 0.39932746, "learning_rate": 1.22e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88521473, "epoch": 0.78022089, "global_step/max_steps": "1925/2468", "percentage": "78.00%", "elapsed_time": "7h 33m 33s", "remaining_time": "2h 7m 56s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070737} +{"loss": 0.33273926, "grad_norm": 0.37402248, "learning_rate": 1.2e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88825806, "epoch": 0.78224744, "global_step/max_steps": "1930/2468", "percentage": "78.20%", "elapsed_time": "7h 34m 43s", "remaining_time": "2h 6m 45s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070739} +{"loss": 0.34274902, "grad_norm": 0.43087313, "learning_rate": 1.17e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.883641, "epoch": 0.78427399, "global_step/max_steps": "1935/2468", "percentage": "78.40%", "elapsed_time": "7h 35m 51s", "remaining_time": "2h 5m 34s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070744} +{"loss": 0.34677734, "grad_norm": 0.43539366, "learning_rate": 1.15e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88371741, "epoch": 0.78630054, "global_step/max_steps": "1940/2468", "percentage": "78.61%", "elapsed_time": "7h 36m 59s", "remaining_time": "2h 4m 22s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070752} +{"loss": 0.31210938, "grad_norm": 0.40453237, "learning_rate": 1.13e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89604244, "epoch": 0.78832708, "global_step/max_steps": "1945/2468", "percentage": "78.81%", "elapsed_time": "7h 38m 10s", "remaining_time": "2h 3m 12s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070752} +{"loss": 0.34001465, "grad_norm": 0.46798185, "learning_rate": 1.11e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88427692, "epoch": 0.79035363, "global_step/max_steps": "1950/2468", "percentage": "79.01%", "elapsed_time": "7h 39m 18s", "remaining_time": "2h 2m 0s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070759} +{"loss": 0.33603516, "grad_norm": 0.51827979, "learning_rate": 1.09e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88824366, "epoch": 0.79238018, "global_step/max_steps": "1955/2468", "percentage": "79.21%", "elapsed_time": "7h 40m 29s", "remaining_time": "2h 0m 50s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070757} +{"loss": 0.34899902, "grad_norm": 0.40419793, "learning_rate": 1.07e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88244556, "epoch": 0.79440673, "global_step/max_steps": "1960/2468", "percentage": "79.42%", "elapsed_time": "7h 41m 37s", "remaining_time": "1h 59m 38s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070765} +{"loss": 0.35170898, "grad_norm": 0.47690392, "learning_rate": 1.05e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8830511, "epoch": 0.79643328, "global_step/max_steps": "1965/2468", "percentage": "79.62%", "elapsed_time": "7h 42m 44s", "remaining_time": "1h 58m 27s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070773} +{"loss": 0.30078125, "grad_norm": 0.35004881, "learning_rate": 1.03e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.90151325, "epoch": 0.79845982, "global_step/max_steps": "1970/2468", "percentage": "79.82%", "elapsed_time": "7h 43m 55s", "remaining_time": "1h 57m 16s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070772} +{"loss": 0.33249512, "grad_norm": 0.43832195, "learning_rate": 1.01e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88758562, "epoch": 0.80048637, "global_step/max_steps": "1975/2468", "percentage": "80.02%", "elapsed_time": "7h 45m 2s", "remaining_time": "1h 56m 4s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070783} +{"loss": 0.32746582, "grad_norm": 0.33005771, "learning_rate": 9.9e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89008616, "epoch": 0.80251292, "global_step/max_steps": "1980/2468", "percentage": "80.23%", "elapsed_time": "7h 46m 11s", "remaining_time": "1h 54m 54s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070785} +{"loss": 0.31999512, "grad_norm": 0.26178536, "learning_rate": 9.7e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89581067, "epoch": 0.80453947, "global_step/max_steps": "1985/2468", "percentage": "80.43%", "elapsed_time": "7h 47m 22s", "remaining_time": "1h 53m 43s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070785} +{"loss": 0.35065918, "grad_norm": 0.55027616, "learning_rate": 9.5e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88090435, "epoch": 0.80656601, "global_step/max_steps": "1990/2468", "percentage": "80.63%", "elapsed_time": "7h 48m 29s", "remaining_time": "1h 52m 31s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070794} +{"loss": 0.34377441, "grad_norm": 0.42342988, "learning_rate": 9.3e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8829914, "epoch": 0.80859256, "global_step/max_steps": "1995/2468", "percentage": "80.83%", "elapsed_time": "7h 49m 37s", "remaining_time": "1h 51m 20s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.0708} +{"loss": 0.34118652, "grad_norm": 0.40228805, "learning_rate": 9.1e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8843093, "epoch": 0.81061911, "global_step/max_steps": "2000/2468", "percentage": "81.04%", "elapsed_time": "7h 50m 47s", "remaining_time": "1h 50m 9s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070804} +{"loss": 0.33823242, "grad_norm": 0.41832343, "learning_rate": 9e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88554041, "epoch": 0.81264566, "global_step/max_steps": "2005/2468", "percentage": "81.24%", "elapsed_time": "7h 51m 55s", "remaining_time": "1h 48m 58s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07081} +{"loss": 0.32199707, "grad_norm": 0.40920776, "learning_rate": 8.8e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89117681, "epoch": 0.81467221, "global_step/max_steps": "2010/2468", "percentage": "81.44%", "elapsed_time": "7h 53m 5s", "remaining_time": "1h 47m 47s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070811} +{"loss": 0.34108887, "grad_norm": 0.43547738, "learning_rate": 8.6e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88576204, "epoch": 0.81669875, "global_step/max_steps": "2015/2468", "percentage": "81.65%", "elapsed_time": "7h 54m 14s", "remaining_time": "1h 46m 36s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070815} +{"loss": 0.33508301, "grad_norm": 0.49091804, "learning_rate": 8.4e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88649694, "epoch": 0.8187253, "global_step/max_steps": "2020/2468", "percentage": "81.85%", "elapsed_time": "7h 55m 24s", "remaining_time": "1h 45m 26s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070817} +{"loss": 0.32097168, "grad_norm": 0.40985855, "learning_rate": 8.2e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89268544, "epoch": 0.82075185, "global_step/max_steps": "2025/2468", "percentage": "82.05%", "elapsed_time": "7h 56m 35s", "remaining_time": "1h 44m 15s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070816} +{"loss": 0.31489258, "grad_norm": 0.35227603, "learning_rate": 8e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89347643, "epoch": 0.8227784, "global_step/max_steps": "2030/2468", "percentage": "82.25%", "elapsed_time": "7h 57m 47s", "remaining_time": "1h 43m 5s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070811} +{"loss": 0.31398926, "grad_norm": 0.4074994, "learning_rate": 7.9e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8930228, "epoch": 0.82480494, "global_step/max_steps": "2035/2468", "percentage": "82.46%", "elapsed_time": "7h 58m 57s", "remaining_time": "1h 41m 54s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070813} +{"loss": 0.32614746, "grad_norm": 0.40438449, "learning_rate": 7.7e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89052816, "epoch": 0.82683149, "global_step/max_steps": "2040/2468", "percentage": "82.66%", "elapsed_time": "8h 0m 5s", "remaining_time": "1h 40m 43s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070821} +{"loss": 0.33327637, "grad_norm": 0.4920688, "learning_rate": 7.5e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88745879, "epoch": 0.82885804, "global_step/max_steps": "2045/2468", "percentage": "82.86%", "elapsed_time": "8h 1m 12s", "remaining_time": "1h 39m 32s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070828} +{"loss": 0.27731934, "grad_norm": 0.32292449, "learning_rate": 7.3e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.90842331, "epoch": 0.83088459, "global_step/max_steps": "2050/2468", "percentage": "83.06%", "elapsed_time": "8h 2m 24s", "remaining_time": "1h 38m 21s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070825} +{"loss": 0.33156738, "grad_norm": 0.39507738, "learning_rate": 7.2e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88755335, "epoch": 0.83291114, "global_step/max_steps": "2055/2468", "percentage": "83.27%", "elapsed_time": "8h 3m 32s", "remaining_time": "1h 37m 10s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070831} +{"loss": 0.3402832, "grad_norm": 0.43537, "learning_rate": 7e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88509416, "epoch": 0.83493768, "global_step/max_steps": "2060/2468", "percentage": "83.47%", "elapsed_time": "8h 4m 43s", "remaining_time": "1h 36m 0s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07083} +{"loss": 0.3458252, "grad_norm": 0.47620904, "learning_rate": 6.8e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88320074, "epoch": 0.83696423, "global_step/max_steps": "2065/2468", "percentage": "83.67%", "elapsed_time": "8h 5m 51s", "remaining_time": "1h 34m 49s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070838} +{"loss": 0.32021484, "grad_norm": 0.32710177, "learning_rate": 6.7e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89213892, "epoch": 0.83899078, "global_step/max_steps": "2070/2468", "percentage": "83.87%", "elapsed_time": "8h 6m 59s", "remaining_time": "1h 33m 38s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070842} +{"loss": 0.31948242, "grad_norm": 0.36265028, "learning_rate": 6.5e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89159776, "epoch": 0.84101733, "global_step/max_steps": "2075/2468", "percentage": "84.08%", "elapsed_time": "8h 8m 6s", "remaining_time": "1h 32m 26s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070851} +{"loss": 0.34445801, "grad_norm": 0.44921699, "learning_rate": 6.3e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88436873, "epoch": 0.84304387, "global_step/max_steps": "2080/2468", "percentage": "84.28%", "elapsed_time": "8h 9m 14s", "remaining_time": "1h 31m 15s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070858} +{"loss": 0.34221191, "grad_norm": 0.42785132, "learning_rate": 6.2e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88403609, "epoch": 0.84507042, "global_step/max_steps": "2085/2468", "percentage": "84.48%", "elapsed_time": "8h 10m 21s", "remaining_time": "1h 30m 4s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070868} +{"loss": 0.34770508, "grad_norm": 0.44158694, "learning_rate": 6e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88357076, "epoch": 0.84709697, "global_step/max_steps": "2090/2468", "percentage": "84.68%", "elapsed_time": "8h 11m 28s", "remaining_time": "1h 28m 53s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070875} +{"loss": 0.3265625, "grad_norm": 0.37700477, "learning_rate": 5.9e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89003568, "epoch": 0.84912352, "global_step/max_steps": "2095/2468", "percentage": "84.89%", "elapsed_time": "8h 12m 38s", "remaining_time": "1h 27m 42s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070876} +{"loss": 0.33596191, "grad_norm": 0.37200838, "learning_rate": 5.7e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88671937, "epoch": 0.85115007, "global_step/max_steps": "2100/2468", "percentage": "85.09%", "elapsed_time": "8h 13m 47s", "remaining_time": "1h 26m 31s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070881} +{"loss": 0.346875, "grad_norm": 0.47252354, "learning_rate": 5.6e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88405826, "epoch": 0.85317661, "global_step/max_steps": "2105/2468", "percentage": "85.29%", "elapsed_time": "8h 14m 52s", "remaining_time": "1h 25m 20s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070893} +{"loss": 0.32822266, "grad_norm": 0.38058013, "learning_rate": 5.4e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88868917, "epoch": 0.85520316, "global_step/max_steps": "2110/2468", "percentage": "85.49%", "elapsed_time": "8h 16m 1s", "remaining_time": "1h 24m 9s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070897} +{"loss": 0.33684082, "grad_norm": 0.44028646, "learning_rate": 5.3e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88588099, "epoch": 0.85722971, "global_step/max_steps": "2115/2468", "percentage": "85.70%", "elapsed_time": "8h 17m 12s", "remaining_time": "1h 22m 59s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070897} +{"loss": 0.32897949, "grad_norm": 0.36741778, "learning_rate": 5.1e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88828737, "epoch": 0.85925626, "global_step/max_steps": "2120/2468", "percentage": "85.90%", "elapsed_time": "8h 18m 17s", "remaining_time": "1h 21m 47s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070908} +{"loss": 0.34401855, "grad_norm": 0.44213435, "learning_rate": 5e-07, "tIoU": 0.99893551, "tRecall_30": 0.99977684, "tRecall_50": 0.99977684, "tRecall_70": 0.99977684, "token_acc": 0.88393752, "epoch": 0.8612828, "global_step/max_steps": "2125/2468", "percentage": "86.10%", "elapsed_time": "8h 19m 24s", "remaining_time": "1h 20m 36s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070918} +{"loss": 0.3519043, "grad_norm": 0.43405816, "learning_rate": 4.8e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88181601, "epoch": 0.86330935, "global_step/max_steps": "2130/2468", "percentage": "86.30%", "elapsed_time": "8h 20m 29s", "remaining_time": "1h 19m 25s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07093} +{"loss": 0.30791016, "grad_norm": 0.37876493, "learning_rate": 4.7e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89612733, "epoch": 0.8653359, "global_step/max_steps": "2135/2468", "percentage": "86.51%", "elapsed_time": "8h 21m 39s", "remaining_time": "1h 18m 14s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070932} +{"loss": 0.3364502, "grad_norm": 0.42582822, "learning_rate": 4.6e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88638915, "epoch": 0.86736245, "global_step/max_steps": "2140/2468", "percentage": "86.71%", "elapsed_time": "8h 22m 49s", "remaining_time": "1h 17m 4s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070932} +{"loss": 0.34580078, "grad_norm": 0.45081168, "learning_rate": 4.4e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88345134, "epoch": 0.869389, "global_step/max_steps": "2145/2468", "percentage": "86.91%", "elapsed_time": "8h 23m 57s", "remaining_time": "1h 15m 53s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070937} +{"loss": 0.30915527, "grad_norm": 0.34580702, "learning_rate": 4.3e-07, "tIoU": 0.99999647, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8950528, "epoch": 0.87141554, "global_step/max_steps": "2150/2468", "percentage": "87.12%", "elapsed_time": "8h 25m 11s", "remaining_time": "1h 14m 43s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070931} +{"loss": 0.35849609, "grad_norm": 0.53368694, "learning_rate": 4.2e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8784318, "epoch": 0.87344209, "global_step/max_steps": "2155/2468", "percentage": "87.32%", "elapsed_time": "8h 26m 20s", "remaining_time": "1h 13m 32s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070933} +{"loss": 0.33815918, "grad_norm": 0.35457611, "learning_rate": 4e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88624577, "epoch": 0.87546864, "global_step/max_steps": "2160/2468", "percentage": "87.52%", "elapsed_time": "8h 27m 29s", "remaining_time": "1h 12m 21s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070937} +{"loss": 0.31135254, "grad_norm": 0.37431887, "learning_rate": 3.9e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89357439, "epoch": 0.87749519, "global_step/max_steps": "2165/2468", "percentage": "87.72%", "elapsed_time": "8h 28m 41s", "remaining_time": "1h 11m 11s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070933} +{"loss": 0.35351562, "grad_norm": 0.44529307, "learning_rate": 3.8e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88123604, "epoch": 0.87952173, "global_step/max_steps": "2170/2468", "percentage": "87.93%", "elapsed_time": "8h 29m 49s", "remaining_time": "1h 10m 0s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07094} +{"loss": 0.32648926, "grad_norm": 0.41940665, "learning_rate": 3.7e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88915491, "epoch": 0.88154828, "global_step/max_steps": "2175/2468", "percentage": "88.13%", "elapsed_time": "8h 30m 59s", "remaining_time": "1h 8m 50s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070942} +{"loss": 0.35056152, "grad_norm": 0.40358305, "learning_rate": 3.5e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8810428, "epoch": 0.88357483, "global_step/max_steps": "2180/2468", "percentage": "88.33%", "elapsed_time": "8h 32m 6s", "remaining_time": "1h 7m 39s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070947} +{"loss": 0.34584961, "grad_norm": 0.37540069, "learning_rate": 3.4e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88221577, "epoch": 0.88560138, "global_step/max_steps": "2185/2468", "percentage": "88.53%", "elapsed_time": "8h 33m 14s", "remaining_time": "1h 6m 28s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070954} +{"loss": 0.33671875, "grad_norm": 0.4143151, "learning_rate": 3.3e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88630028, "epoch": 0.88762793, "global_step/max_steps": "2190/2468", "percentage": "88.74%", "elapsed_time": "8h 34m 21s", "remaining_time": "1h 5m 17s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070962} +{"loss": 0.33366699, "grad_norm": 0.44223762, "learning_rate": 3.2e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88712161, "epoch": 0.88965447, "global_step/max_steps": "2195/2468", "percentage": "88.94%", "elapsed_time": "8h 35m 30s", "remaining_time": "1h 4m 6s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070966} +{"loss": 0.31289062, "grad_norm": 0.34741598, "learning_rate": 3.1e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89456626, "epoch": 0.89168102, "global_step/max_steps": "2200/2468", "percentage": "89.14%", "elapsed_time": "8h 36m 39s", "remaining_time": "1h 2m 56s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07097} +{"loss": 0.32885742, "grad_norm": 0.31624061, "learning_rate": 3e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88874552, "epoch": 0.89370757, "global_step/max_steps": "2205/2468", "percentage": "89.34%", "elapsed_time": "8h 37m 47s", "remaining_time": "1h 1m 45s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070974} +{"loss": 0.34553223, "grad_norm": 0.43610123, "learning_rate": 2.8e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88553563, "epoch": 0.89573412, "global_step/max_steps": "2210/2468", "percentage": "89.55%", "elapsed_time": "8h 38m 57s", "remaining_time": "1h 0m 35s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070976} +{"loss": 0.33029785, "grad_norm": 0.4188253, "learning_rate": 2.7e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88849118, "epoch": 0.89776066, "global_step/max_steps": "2215/2468", "percentage": "89.75%", "elapsed_time": "8h 40m 6s", "remaining_time": "59m 24s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070978} +{"loss": 0.32077637, "grad_norm": 0.48173168, "learning_rate": 2.6e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89126999, "epoch": 0.89978721, "global_step/max_steps": "2220/2468", "percentage": "89.95%", "elapsed_time": "8h 41m 17s", "remaining_time": "58m 14s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070977} +{"loss": 0.33603516, "grad_norm": 0.3567912, "learning_rate": 2.5e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.886389, "epoch": 0.90181376, "global_step/max_steps": "2225/2468", "percentage": "90.15%", "elapsed_time": "8h 42m 27s", "remaining_time": "57m 3s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070978} +{"loss": 0.30375977, "grad_norm": 0.42567268, "learning_rate": 2.4e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89727138, "epoch": 0.90384031, "global_step/max_steps": "2230/2468", "percentage": "90.36%", "elapsed_time": "8h 43m 40s", "remaining_time": "55m 53s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070973} +{"loss": 0.34375, "grad_norm": 0.4955014, "learning_rate": 2.3e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88584391, "epoch": 0.90586686, "global_step/max_steps": "2235/2468", "percentage": "90.56%", "elapsed_time": "8h 44m 51s", "remaining_time": "54m 43s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070972} +{"loss": 0.34416504, "grad_norm": 0.38924527, "learning_rate": 2.2e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88434897, "epoch": 0.9078934, "global_step/max_steps": "2240/2468", "percentage": "90.76%", "elapsed_time": "8h 46m 2s", "remaining_time": "53m 32s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07097} +{"loss": 0.32607422, "grad_norm": 0.38762596, "learning_rate": 2.1e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88822613, "epoch": 0.90991995, "global_step/max_steps": "2245/2468", "percentage": "90.96%", "elapsed_time": "8h 47m 11s", "remaining_time": "52m 21s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070974} +{"loss": 0.33171387, "grad_norm": 0.52062219, "learning_rate": 2e-07, "tIoU": 0.99821224, "tRecall_30": 0.99960938, "tRecall_50": 0.99960938, "tRecall_70": 0.99960938, "token_acc": 0.88741093, "epoch": 0.9119465, "global_step/max_steps": "2250/2468", "percentage": "91.17%", "elapsed_time": "8h 48m 20s", "remaining_time": "51m 11s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070978} +{"loss": 0.34367676, "grad_norm": 0.59368873, "learning_rate": 1.9e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88382483, "epoch": 0.91397305, "global_step/max_steps": "2255/2468", "percentage": "91.37%", "elapsed_time": "8h 49m 26s", "remaining_time": "50m 0s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070987} +{"loss": 0.34084473, "grad_norm": 0.52687538, "learning_rate": 1.9e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88600137, "epoch": 0.91599959, "global_step/max_steps": "2260/2468", "percentage": "91.57%", "elapsed_time": "8h 50m 34s", "remaining_time": "48m 49s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070991} +{"loss": 0.3270752, "grad_norm": 0.44447538, "learning_rate": 1.8e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89318273, "epoch": 0.91802614, "global_step/max_steps": "2265/2468", "percentage": "91.77%", "elapsed_time": "8h 51m 46s", "remaining_time": "47m 39s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070988} +{"loss": 0.32766113, "grad_norm": 0.34051868, "learning_rate": 1.7e-07, "tIoU": 0.99994564, "tRecall_30": 0.99994316, "tRecall_50": 0.99994316, "tRecall_70": 0.99994316, "token_acc": 0.89062077, "epoch": 0.92005269, "global_step/max_steps": "2270/2468", "percentage": "91.98%", "elapsed_time": "8h 52m 57s", "remaining_time": "46m 29s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070987} +{"loss": 0.34050293, "grad_norm": 0.40930411, "learning_rate": 1.6e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88468963, "epoch": 0.92207924, "global_step/max_steps": "2275/2468", "percentage": "92.18%", "elapsed_time": "8h 54m 8s", "remaining_time": "45m 18s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070987} +{"loss": 0.31638184, "grad_norm": 0.35660627, "learning_rate": 1.5e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89336543, "epoch": 0.92410579, "global_step/max_steps": "2280/2468", "percentage": "92.38%", "elapsed_time": "8h 55m 19s", "remaining_time": "44m 8s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070984} +{"loss": 0.34787598, "grad_norm": 0.48520005, "learning_rate": 1.4e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88174911, "epoch": 0.92613233, "global_step/max_steps": "2285/2468", "percentage": "92.59%", "elapsed_time": "8h 56m 29s", "remaining_time": "42m 57s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070986} +{"loss": 0.3416748, "grad_norm": 0.49744278, "learning_rate": 1.4e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88287119, "epoch": 0.92815888, "global_step/max_steps": "2290/2468", "percentage": "92.79%", "elapsed_time": "8h 57m 37s", "remaining_time": "41m 47s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070992} +{"loss": 0.31433105, "grad_norm": 0.41715986, "learning_rate": 1.3e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89348461, "epoch": 0.93018543, "global_step/max_steps": "2295/2468", "percentage": "92.99%", "elapsed_time": "8h 58m 48s", "remaining_time": "40m 36s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07099} +{"loss": 0.3171875, "grad_norm": 0.32260588, "learning_rate": 1.2e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89430947, "epoch": 0.93221198, "global_step/max_steps": "2300/2468", "percentage": "93.19%", "elapsed_time": "8h 59m 59s", "remaining_time": "39m 26s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070988} +{"loss": 0.31831055, "grad_norm": 0.44351634, "learning_rate": 1.1e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89232106, "epoch": 0.93423852, "global_step/max_steps": "2305/2468", "percentage": "93.40%", "elapsed_time": "9h 1m 8s", "remaining_time": "38m 16s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070991} +{"loss": 0.33178711, "grad_norm": 0.45417944, "learning_rate": 1.1e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88951403, "epoch": 0.93626507, "global_step/max_steps": "2310/2468", "percentage": "93.60%", "elapsed_time": "9h 2m 26s", "remaining_time": "37m 6s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070976} +{"loss": 0.33115234, "grad_norm": 0.45320377, "learning_rate": 1e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8877488, "epoch": 0.93829162, "global_step/max_steps": "2315/2468", "percentage": "93.80%", "elapsed_time": "9h 3m 34s", "remaining_time": "35m 55s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07098} +{"loss": 0.31811523, "grad_norm": 0.43640557, "learning_rate": 9e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89412209, "epoch": 0.94031817, "global_step/max_steps": "2320/2468", "percentage": "94.00%", "elapsed_time": "9h 4m 45s", "remaining_time": "34m 45s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070979} +{"loss": 0.32717285, "grad_norm": 0.39802286, "learning_rate": 9e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89033468, "epoch": 0.94234472, "global_step/max_steps": "2325/2468", "percentage": "94.21%", "elapsed_time": "9h 5m 55s", "remaining_time": "33m 34s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070981} +{"loss": 0.31474609, "grad_norm": 0.35078153, "learning_rate": 8e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89402823, "epoch": 0.94437126, "global_step/max_steps": "2330/2468", "percentage": "94.41%", "elapsed_time": "9h 7m 3s", "remaining_time": "32m 24s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070985} +{"loss": 0.33984375, "grad_norm": 0.42213422, "learning_rate": 8e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8851875, "epoch": 0.94639781, "global_step/max_steps": "2335/2468", "percentage": "94.61%", "elapsed_time": "9h 8m 11s", "remaining_time": "31m 13s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070991} +{"loss": 0.3451416, "grad_norm": 0.46918657, "learning_rate": 7e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88440889, "epoch": 0.94842436, "global_step/max_steps": "2340/2468", "percentage": "94.81%", "elapsed_time": "9h 9m 19s", "remaining_time": "30m 2s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070996} +{"loss": 0.33742676, "grad_norm": 0.26639035, "learning_rate": 7e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88784928, "epoch": 0.95045091, "global_step/max_steps": "2345/2468", "percentage": "95.02%", "elapsed_time": "9h 10m 28s", "remaining_time": "28m 52s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.071} +{"loss": 0.31958008, "grad_norm": 0.43997228, "learning_rate": 6e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89262504, "epoch": 0.95247745, "global_step/max_steps": "2350/2468", "percentage": "95.22%", "elapsed_time": "9h 11m 37s", "remaining_time": "27m 41s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.071002} +{"loss": 0.33017578, "grad_norm": 0.37927336, "learning_rate": 5e-08, "tIoU": 0.99986496, "tRecall_30": 0.99984379, "tRecall_50": 0.99984379, "tRecall_70": 0.99984379, "token_acc": 0.88976174, "epoch": 0.954504, "global_step/max_steps": "2355/2468", "percentage": "95.42%", "elapsed_time": "9h 12m 46s", "remaining_time": "26m 31s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.071005} +{"loss": 0.3291748, "grad_norm": 0.41160727, "learning_rate": 5e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88863117, "epoch": 0.95653055, "global_step/max_steps": "2360/2468", "percentage": "95.62%", "elapsed_time": "9h 13m 54s", "remaining_time": "25m 20s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.071011} +{"loss": 0.34379883, "grad_norm": 0.40575126, "learning_rate": 5e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88511022, "epoch": 0.9585571, "global_step/max_steps": "2365/2468", "percentage": "95.83%", "elapsed_time": "9h 15m 1s", "remaining_time": "24m 10s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.071018} +{"loss": 0.32990723, "grad_norm": 0.45322505, "learning_rate": 4e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88913112, "epoch": 0.96058365, "global_step/max_steps": "2370/2468", "percentage": "96.03%", "elapsed_time": "9h 16m 10s", "remaining_time": "22m 59s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.071021} +{"loss": 0.31518555, "grad_norm": 0.37385947, "learning_rate": 4e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89295977, "epoch": 0.96261019, "global_step/max_steps": "2375/2468", "percentage": "96.23%", "elapsed_time": "9h 17m 59s", "remaining_time": "21m 51s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070938} +{"loss": 0.3309082, "grad_norm": 0.39025664, "learning_rate": 3e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88768506, "epoch": 0.96463674, "global_step/max_steps": "2380/2468", "percentage": "96.43%", "elapsed_time": "9h 20m 3s", "remaining_time": "20m 42s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070826} +{"loss": 0.32819824, "grad_norm": 0.49842587, "learning_rate": 3e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89276055, "epoch": 0.96666329, "global_step/max_steps": "2385/2468", "percentage": "96.64%", "elapsed_time": "9h 22m 11s", "remaining_time": "19m 33s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070705} +{"loss": 0.31704102, "grad_norm": 0.35348463, "learning_rate": 3e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89247716, "epoch": 0.96868984, "global_step/max_steps": "2390/2468", "percentage": "96.84%", "elapsed_time": "9h 24m 17s", "remaining_time": "18m 24s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07059} +{"loss": 0.34553223, "grad_norm": 0.40492561, "learning_rate": 2e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88431159, "epoch": 0.97071638, "global_step/max_steps": "2395/2468", "percentage": "97.04%", "elapsed_time": "9h 26m 22s", "remaining_time": "17m 15s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070478} +{"loss": 0.3220459, "grad_norm": 0.37525889, "learning_rate": 2e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89011735, "epoch": 0.97274293, "global_step/max_steps": "2400/2468", "percentage": "97.24%", "elapsed_time": "9h 28m 30s", "remaining_time": "16m 6s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070359} +{"loss": 0.32915039, "grad_norm": 0.45800465, "learning_rate": 2e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88856642, "epoch": 0.97476948, "global_step/max_steps": "2405/2468", "percentage": "97.45%", "elapsed_time": "9h 30m 32s", "remaining_time": "14m 56s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070254} +{"loss": 0.32165527, "grad_norm": 0.35217658, "learning_rate": 1e-08, "tIoU": 0.99994259, "tRecall_30": 0.99994316, "tRecall_50": 0.99994316, "tRecall_70": 0.99994316, "token_acc": 0.89333333, "epoch": 0.97679603, "global_step/max_steps": "2410/2468", "percentage": "97.65%", "elapsed_time": "9h 32m 39s", "remaining_time": "13m 46s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.07014} +{"loss": 0.30314941, "grad_norm": 0.4140844, "learning_rate": 1e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89875337, "epoch": 0.97882258, "global_step/max_steps": "2415/2468", "percentage": "97.85%", "elapsed_time": "9h 34m 50s", "remaining_time": "12m 36s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.070019} +{"loss": 0.33303223, "grad_norm": 0.34325486, "learning_rate": 1e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88785364, "epoch": 0.98084912, "global_step/max_steps": "2420/2468", "percentage": "98.06%", "elapsed_time": "9h 36m 54s", "remaining_time": "11m 26s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069913} +{"loss": 0.31040039, "grad_norm": 0.37910819, "learning_rate": 1e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89476837, "epoch": 0.98287567, "global_step/max_steps": "2425/2468", "percentage": "98.26%", "elapsed_time": "9h 39m 1s", "remaining_time": "10m 16s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069801} +{"loss": 0.3291748, "grad_norm": 0.46881887, "learning_rate": 1e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88971824, "epoch": 0.98490222, "global_step/max_steps": "2430/2468", "percentage": "98.46%", "elapsed_time": "9h 41m 6s", "remaining_time": "9m 5s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069694} +{"loss": 0.33132324, "grad_norm": 0.3836017, "learning_rate": 0.0, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8881468, "epoch": 0.98692877, "global_step/max_steps": "2435/2468", "percentage": "98.66%", "elapsed_time": "9h 43m 9s", "remaining_time": "7m 54s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069593} +{"loss": 0.34050293, "grad_norm": 0.53583324, "learning_rate": 0.0, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8851056, "epoch": 0.98895531, "global_step/max_steps": "2440/2468", "percentage": "98.87%", "elapsed_time": "9h 45m 11s", "remaining_time": "6m 42s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069492} +{"loss": 0.33144531, "grad_norm": 0.331191, "learning_rate": 0.0, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88887365, "epoch": 0.99098186, "global_step/max_steps": "2445/2468", "percentage": "99.07%", "elapsed_time": "9h 47m 15s", "remaining_time": "5m 31s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069389} +{"loss": 0.31621094, "grad_norm": 0.42428568, "learning_rate": 0.0, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89250403, "epoch": 0.99300841, "global_step/max_steps": "2450/2468", "percentage": "99.27%", "elapsed_time": "9h 49m 24s", "remaining_time": "4m 19s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069279} +{"loss": 0.33442383, "grad_norm": 0.49177971, "learning_rate": 0.0, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88678676, "epoch": 0.99503496, "global_step/max_steps": "2455/2468", "percentage": "99.47%", "elapsed_time": "9h 51m 31s", "remaining_time": "3m 7s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069172} +{"loss": 0.3192627, "grad_norm": 0.50943655, "learning_rate": 0.0, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.89227916, "epoch": 0.99706151, "global_step/max_steps": "2460/2468", "percentage": "99.68%", "elapsed_time": "9h 53m 42s", "remaining_time": "1m 55s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.069058} +{"loss": 0.36032715, "grad_norm": 0.36331922, "learning_rate": 0.0, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.87848637, "epoch": 0.99908805, "global_step/max_steps": "2465/2468", "percentage": "99.88%", "elapsed_time": "9h 55m 49s", "remaining_time": "43s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068953} +{"train_runtime": 35960.2414, "train_samples_per_second": 2.196, "train_steps_per_second": 0.069, "total_flos": 8.412078391672858e+19, "train_loss": 0.36147493, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88688308, "epoch": 1.0, "global_step/max_steps": "2468/2468", "percentage": "100.00%", "elapsed_time": "9h 59m 17s", "remaining_time": "0s", "memory(GiB)": 55.82, "train_speed(iter/s)": 0.068637} +{"model_parameter_info": "Qwen3VLForConditionalGeneration: 4437.8158M Params (4131.5732M Trainable [93.0993%]), 0.0001M Buffers.", "last_model_checkpoint": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot/v1-20260310-063639/checkpoint-2468", "best_model_checkpoint": null, "best_metric": null, "global_step": 2468, "log_history": [{"loss": 0.703125, "grad_norm": 1.2059801816940308, "learning_rate": 1.3333333333333336e-07, "tIoU": 0.2516516149044037, "tRecall_30": 0.7859375476837158, "tRecall_50": 0.5697917342185974, "tRecall_70": 0.5463542342185974, "token_acc": 0.8105131343028716, "epoch": 0.0004053095551727632, "step": 1}, {"loss": 0.733642578125, "grad_norm": 1.4883641004562378, "learning_rate": 6.666666666666667e-07, "tIoU": 0.2318701297044754, "tRecall_30": 0.7443870902061462, "tRecall_50": 0.6183698773384094, "tRecall_70": 0.5276147723197937, "token_acc": 0.8065031164604378, "epoch": 0.002026547775863816, "step": 5}, {"loss": 0.74443359375, "grad_norm": 1.1400638818740845, "learning_rate": 1.3333333333333334e-06, "tIoU": 0.252414608001709, "tRecall_30": 0.7867266654968261, "tRecall_50": 0.6461015701293945, "tRecall_70": 0.5109973907470703, "token_acc": 0.8026641380519708, "epoch": 0.004053095551727632, "step": 10}, {"loss": 0.675390625, "grad_norm": 0.8875113129615784, "learning_rate": 2.0000000000000003e-06, "tIoU": 0.2845264196395874, "tRecall_30": 0.8465773582458496, "tRecall_50": 0.7617856979370117, "tRecall_70": 0.7227232456207275, "token_acc": 0.8118741947509205, "epoch": 0.006079643327591448, "step": 15}, {"loss": 0.573974609375, "grad_norm": 1.0106897354125977, "learning_rate": 2.666666666666667e-06, "tIoU": 0.24748537540435792, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8294778755378588, "epoch": 0.008106191103455264, "step": 20}, {"loss": 0.474853515625, "grad_norm": 0.7312051057815552, "learning_rate": 3.3333333333333333e-06, "tIoU": 0.8458622932434082, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.852569836141382, "epoch": 0.01013273887931908, "step": 25}, {"loss": 0.422412109375, "grad_norm": 0.48748230934143066, "learning_rate": 4.000000000000001e-06, "tIoU": 0.9924302101135254, "tRecall_30": 0.9893749237060547, "tRecall_50": 0.9893749237060547, "tRecall_70": 0.9893749237060547, "token_acc": 0.8628151138194027, "epoch": 0.012159286655182896, "step": 30}, {"loss": 0.4025146484375, "grad_norm": 0.5568686723709106, "learning_rate": 4.666666666666667e-06, "tIoU": 0.9956202507019043, "tRecall_30": 0.9938838958740235, "tRecall_50": 0.9938838958740235, "tRecall_70": 0.9938838958740235, "token_acc": 0.8690114445460885, "epoch": 0.014185834431046712, "step": 35}, {"loss": 0.4341064453125, "grad_norm": 0.6236920356750488, "learning_rate": 5.333333333333334e-06, "tIoU": 0.9995620727539063, "tRecall_30": 0.9987499237060546, "tRecall_50": 0.9987499237060546, "tRecall_70": 0.9987499237060546, "token_acc": 0.8577157247139078, "epoch": 0.016212382206910528, "step": 40}, {"loss": 0.3864501953125, "grad_norm": 0.6035606861114502, "learning_rate": 6e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8727699530516432, "epoch": 0.018238929982774345, "step": 45}, {"loss": 0.3807373046875, "grad_norm": 0.4480346739292145, "learning_rate": 6.666666666666667e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8747324914642315, "epoch": 0.02026547775863816, "step": 50}, {"loss": 0.420947265625, "grad_norm": 0.5582020282745361, "learning_rate": 7.333333333333333e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8620249574588769, "epoch": 0.022292025534501977, "step": 55}, {"loss": 0.3962890625, "grad_norm": 0.6299155950546265, "learning_rate": 8.000000000000001e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8698634443214969, "epoch": 0.02431857331036579, "step": 60}, {"loss": 0.361181640625, "grad_norm": 0.3295726478099823, "learning_rate": 8.666666666666668e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8810466902481946, "epoch": 0.02634512108622961, "step": 65}, {"loss": 0.42099609375, "grad_norm": 0.5453260540962219, "learning_rate": 9.333333333333334e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8612275187707351, "epoch": 0.028371668862093424, "step": 70}, {"loss": 0.3656982421875, "grad_norm": 0.3517334461212158, "learning_rate": 1e-05, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8824227347762289, "epoch": 0.03039821663795724, "step": 75}, {"loss": 0.405224609375, "grad_norm": 0.6326484084129333, "learning_rate": 9.999892280877701e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.866988653320259, "epoch": 0.032424764413821056, "step": 80}, {"loss": 0.374755859375, "grad_norm": 0.4170985221862793, "learning_rate": 9.999569128152163e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8817892842977771, "epoch": 0.03445131218968487, "step": 85}, {"loss": 0.3975830078125, "grad_norm": 0.6364180445671082, "learning_rate": 9.99903055574728e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8686657147390132, "epoch": 0.03647785996554869, "step": 90}, {"loss": 0.386767578125, "grad_norm": 0.42150235176086426, "learning_rate": 9.99827658686887e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.876250368819027, "epoch": 0.038504407741412505, "step": 95}, {"loss": 0.3949951171875, "grad_norm": 0.550065279006958, "learning_rate": 9.997307254003679e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8704353888079093, "epoch": 0.04053095551727632, "step": 100}, {"loss": 0.3871826171875, "grad_norm": 0.419331431388855, "learning_rate": 9.99612259891798e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8742890360525686, "epoch": 0.042557503293140134, "step": 105}, {"loss": 0.396240234375, "grad_norm": 0.4708976447582245, "learning_rate": 9.99472267265578e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8678600939563964, "epoch": 0.044584051069003955, "step": 110}, {"loss": 0.392578125, "grad_norm": 0.4568882882595062, "learning_rate": 9.993107535536606e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8693056473022323, "epoch": 0.04661059884486777, "step": 115}, {"loss": 0.375634765625, "grad_norm": 0.46889740228652954, "learning_rate": 9.99127725715292e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8766478017806612, "epoch": 0.04863714662073158, "step": 120}, {"loss": 0.379248046875, "grad_norm": 0.4288806617259979, "learning_rate": 9.989231916367114e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8759755899858728, "epoch": 0.0506636943965954, "step": 125}, {"loss": 0.4231689453125, "grad_norm": 0.460036039352417, "learning_rate": 9.986971601308115e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8616542750929368, "epoch": 0.05269024217245922, "step": 130}, {"loss": 0.36162109375, "grad_norm": 0.5301831364631653, "learning_rate": 9.984496409367586e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8786379884342778, "epoch": 0.05471678994832303, "step": 135}, {"loss": 0.4042724609375, "grad_norm": 0.47949114441871643, "learning_rate": 9.981806447195723e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8658658136974522, "epoch": 0.05674333772418685, "step": 140}, {"loss": 0.368505859375, "grad_norm": 0.4840526282787323, "learning_rate": 9.978901830696678e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8775575871141142, "epoch": 0.05876988550005066, "step": 145}, {"loss": 0.3725830078125, "grad_norm": 0.40476393699645996, "learning_rate": 9.975782685023543e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.877293148093681, "epoch": 0.06079643327591448, "step": 150}, {"loss": 0.37080078125, "grad_norm": 0.425466388463974, "learning_rate": 9.972449144572973e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8802019781715128, "epoch": 0.06282298105177829, "step": 155}, {"loss": 0.4077880859375, "grad_norm": 0.560032069683075, "learning_rate": 9.968901352979388e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8662045556516481, "epoch": 0.06484952882764211, "step": 160}, {"loss": 0.3838623046875, "grad_norm": 0.37908798456192017, "learning_rate": 9.965139463108788e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8728647211558769, "epoch": 0.06687607660350593, "step": 165}, {"loss": 0.4210205078125, "grad_norm": 0.49446025490760803, "learning_rate": 9.961163637052161e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.860889583704378, "epoch": 0.06890262437936974, "step": 170}, {"loss": 0.377734375, "grad_norm": 0.48118704557418823, "learning_rate": 9.956974046118506e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8757897836007655, "epoch": 0.07092917215523356, "step": 175}, {"loss": 0.413671875, "grad_norm": 0.5156424045562744, "learning_rate": 9.952570870827444e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8644025372652602, "epoch": 0.07295571993109738, "step": 180}, {"loss": 0.38076171875, "grad_norm": 0.37811586260795593, "learning_rate": 9.94795430090145e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8748112623583383, "epoch": 0.07498226770696119, "step": 185}, {"loss": 0.3935302734375, "grad_norm": 0.42654335498809814, "learning_rate": 9.943124535257665e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8720547928555367, "epoch": 0.07700881548282501, "step": 190}, {"loss": 0.3921142578125, "grad_norm": 0.43498939275741577, "learning_rate": 9.938081781999337e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8706406149503881, "epoch": 0.07903536325868882, "step": 195}, {"loss": 0.37509765625, "grad_norm": 0.4303409159183502, "learning_rate": 9.932826258406847e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8761669059603685, "epoch": 0.08106191103455264, "step": 200}, {"loss": 0.391943359375, "grad_norm": 0.5523289442062378, "learning_rate": 9.927358190928349e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8704423086005763, "epoch": 0.08308845881041646, "step": 205}, {"loss": 0.399560546875, "grad_norm": 0.4943441152572632, "learning_rate": 9.921677815170018e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8697048980833587, "epoch": 0.08511500658628027, "step": 210}, {"loss": 0.4166015625, "grad_norm": 0.506697416305542, "learning_rate": 9.91578537588589e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8614923144262211, "epoch": 0.08714155436214409, "step": 215}, {"loss": 0.39150390625, "grad_norm": 0.43591195344924927, "learning_rate": 9.909681126967317e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8727103654049425, "epoch": 0.08916810213800791, "step": 220}, {"loss": 0.383642578125, "grad_norm": 0.5096474289894104, "learning_rate": 9.903365331432037e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8732293534195278, "epoch": 0.09119464991387172, "step": 225}, {"loss": 0.422021484375, "grad_norm": 0.4196990132331848, "learning_rate": 9.896838261412828e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8624010501078235, "epoch": 0.09322119768973554, "step": 230}, {"loss": 0.3818115234375, "grad_norm": 0.5218161940574646, "learning_rate": 9.890100198145793e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8751534205991717, "epoch": 0.09524774546559935, "step": 235}, {"loss": 0.392919921875, "grad_norm": 0.37742888927459717, "learning_rate": 9.883151431958236e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8700355812097611, "epoch": 0.09727429324146317, "step": 240}, {"loss": 0.39609375, "grad_norm": 0.46909821033477783, "learning_rate": 9.875992262256156e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8688558903912973, "epoch": 0.09930084101732699, "step": 245}, {"loss": 0.3753662109375, "grad_norm": 0.6484021544456482, "learning_rate": 9.868622997511341e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8763179237631792, "epoch": 0.1013273887931908, "step": 250}, {"loss": 0.3893798828125, "grad_norm": 0.5185469388961792, "learning_rate": 9.861043955248085e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8709359037257222, "epoch": 0.10335393656905462, "step": 255}, {"loss": 0.401123046875, "grad_norm": 0.48241207003593445, "learning_rate": 9.853255462029501e-06, "tIoU": 0.9969846725463867, "tRecall_30": 0.9979166030883789, "tRecall_50": 0.9979166030883789, "tRecall_70": 0.9979166030883789, "token_acc": 0.86794682422452, "epoch": 0.10538048434491844, "step": 260}, {"loss": 0.4021484375, "grad_norm": 0.4943467676639557, "learning_rate": 9.845257853443448e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8677978955351218, "epoch": 0.10740703212078224, "step": 265}, {"loss": 0.3967041015625, "grad_norm": 0.5944103002548218, "learning_rate": 9.83705147408808e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8712227138360563, "epoch": 0.10943357989664607, "step": 270}, {"loss": 0.368603515625, "grad_norm": 0.4518413245677948, "learning_rate": 9.828636677556987e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8792639007674004, "epoch": 0.11146012767250987, "step": 275}, {"loss": 0.408349609375, "grad_norm": 0.4986145496368408, "learning_rate": 9.820013826423966e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8656716417910447, "epoch": 0.1134866754483737, "step": 280}, {"loss": 0.39873046875, "grad_norm": 0.418460488319397, "learning_rate": 9.811183292227406e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8686417924266272, "epoch": 0.11551322322423752, "step": 285}, {"loss": 0.407666015625, "grad_norm": 0.5732330083847046, "learning_rate": 9.802145455454258e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8654167783778263, "epoch": 0.11753977100010132, "step": 290}, {"loss": 0.392236328125, "grad_norm": 0.42870670557022095, "learning_rate": 9.792900705523661e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8706978203295145, "epoch": 0.11956631877596514, "step": 295}, {"loss": 0.417626953125, "grad_norm": 0.7802377939224243, "learning_rate": 9.783449440770156e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8623496986015365, "epoch": 0.12159286655182897, "step": 300}, {"loss": 0.389306640625, "grad_norm": 0.552323043346405, "learning_rate": 9.773792068426523e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8697585808039884, "epoch": 0.12361941432769277, "step": 305}, {"loss": 0.402001953125, "grad_norm": 0.48368698358535767, "learning_rate": 9.763929004606224e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8682366737867447, "epoch": 0.12564596210355658, "step": 310}, {"loss": 0.37734375, "grad_norm": 0.3376498520374298, "learning_rate": 9.753860674285496e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8767554365694424, "epoch": 0.12767250987942041, "step": 315}, {"loss": 0.376708984375, "grad_norm": 0.26270627975463867, "learning_rate": 9.743587511285017e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8774723504125398, "epoch": 0.12969905765528422, "step": 320}, {"loss": 0.381787109375, "grad_norm": 0.39964982867240906, "learning_rate": 9.733109958251232e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8723410804986917, "epoch": 0.13172560543114803, "step": 325}, {"loss": 0.359912109375, "grad_norm": 0.4558398425579071, "learning_rate": 9.722428466637266e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8835972975966331, "epoch": 0.13375215320701186, "step": 330}, {"loss": 0.417822265625, "grad_norm": 0.6383611559867859, "learning_rate": 9.711543496683477e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8618028783968806, "epoch": 0.13577870098287567, "step": 335}, {"loss": 0.413818359375, "grad_norm": 0.6163831353187561, "learning_rate": 9.700455517397633e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8630407959419495, "epoch": 0.13780524875873948, "step": 340}, {"loss": 0.4010986328125, "grad_norm": 0.5585286021232605, "learning_rate": 9.68916500653469e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8668367878346006, "epoch": 0.13983179653460331, "step": 345}, {"loss": 0.4055908203125, "grad_norm": 0.532497763633728, "learning_rate": 9.677672450576214e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8656914720402236, "epoch": 0.14185834431046712, "step": 350}, {"loss": 0.373193359375, "grad_norm": 0.5201588869094849, "learning_rate": 9.665978344709428e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8762551519158639, "epoch": 0.14388489208633093, "step": 355}, {"loss": 0.39013671875, "grad_norm": 0.39076343178749084, "learning_rate": 9.654083192805854e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8710004305063417, "epoch": 0.14591143986219476, "step": 360}, {"loss": 0.3804931640625, "grad_norm": 0.4675477147102356, "learning_rate": 9.641987507399623e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8750331619439847, "epoch": 0.14793798763805857, "step": 365}, {"loss": 0.3907958984375, "grad_norm": 0.5155381560325623, "learning_rate": 9.629691809665385e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8691968844244972, "epoch": 0.14996453541392238, "step": 370}, {"loss": 0.402197265625, "grad_norm": 0.4530424475669861, "learning_rate": 9.61719662939584e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8673240255194368, "epoch": 0.15199108318978619, "step": 375}, {"loss": 0.39306640625, "grad_norm": 0.4631502330303192, "learning_rate": 9.604502504978934e-06, "tIoU": 0.9999645233154297, "tRecall_30": 1.0, "tRecall_50": 0.9998046875, "tRecall_70": 0.9998046875, "token_acc": 0.8692022617865057, "epoch": 0.15401763096565002, "step": 380}, {"loss": 0.3896240234375, "grad_norm": 0.45243507623672485, "learning_rate": 9.591609983374641e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.871999049987795, "epoch": 0.15604417874151383, "step": 385}, {"loss": 0.3758544921875, "grad_norm": 0.3028954863548279, "learning_rate": 9.578519620091406e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8807784330048143, "epoch": 0.15807072651737764, "step": 390}, {"loss": 0.36181640625, "grad_norm": 0.42049020528793335, "learning_rate": 9.565231979162209e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8822334418785289, "epoch": 0.16009727429324147, "step": 395}, {"loss": 0.3998291015625, "grad_norm": 0.5457665324211121, "learning_rate": 9.551747633120252e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8688331201649054, "epoch": 0.16212382206910528, "step": 400}, {"loss": 0.404541015625, "grad_norm": 0.5921686887741089, "learning_rate": 9.538067162974307e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8673737888832228, "epoch": 0.16415036984496909, "step": 405}, {"loss": 0.4027587890625, "grad_norm": 0.44998884201049805, "learning_rate": 9.524191158183667e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8659621695270872, "epoch": 0.16617691762083292, "step": 410}, {"loss": 0.395751953125, "grad_norm": 0.4771840274333954, "learning_rate": 9.510120216632756e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8684010315925209, "epoch": 0.16820346539669673, "step": 415}, {"loss": 0.3904541015625, "grad_norm": 0.4623829126358032, "learning_rate": 9.495854944605365e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8688424574821375, "epoch": 0.17023001317256053, "step": 420}, {"loss": 0.4014892578125, "grad_norm": 0.38424453139305115, "learning_rate": 9.481395956758524e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8684298649367362, "epoch": 0.17225656094842437, "step": 425}, {"loss": 0.3872802734375, "grad_norm": 0.46710488200187683, "learning_rate": 9.466743876096025e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8724153877983181, "epoch": 0.17428310872428818, "step": 430}, {"loss": 0.3877197265625, "grad_norm": 0.5199359059333801, "learning_rate": 9.451899333941576e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8741814181929807, "epoch": 0.17630965650015198, "step": 435}, {"loss": 0.387841796875, "grad_norm": 0.4269919693470001, "learning_rate": 9.4368629699116e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8710228758766247, "epoch": 0.17833620427601582, "step": 440}, {"loss": 0.37880859375, "grad_norm": 0.4944726526737213, "learning_rate": 9.421635431887669e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8753988453659445, "epoch": 0.18036275205187963, "step": 445}, {"loss": 0.3662109375, "grad_norm": 0.5226631164550781, "learning_rate": 9.406217375988597e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.881288582122199, "epoch": 0.18238929982774343, "step": 450}, {"loss": 0.3919921875, "grad_norm": 0.4460074007511139, "learning_rate": 9.390609466542162e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8687119839296501, "epoch": 0.18441584760360724, "step": 455}, {"loss": 0.395458984375, "grad_norm": 0.41100502014160156, "learning_rate": 9.374812376056486e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8694572158365261, "epoch": 0.18644239537947108, "step": 460}, {"loss": 0.3611083984375, "grad_norm": 0.43790924549102783, "learning_rate": 9.35882678519106e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8846521273125495, "epoch": 0.18846894315533488, "step": 465}, {"loss": 0.3850341796875, "grad_norm": 0.447274386882782, "learning_rate": 9.34265338272741e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8730763461826908, "epoch": 0.1904954909311987, "step": 470}, {"loss": 0.3983154296875, "grad_norm": 0.4919845163822174, "learning_rate": 9.326292865539423e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8671222277848937, "epoch": 0.19252203870706253, "step": 475}, {"loss": 0.37763671875, "grad_norm": 0.4264659881591797, "learning_rate": 9.309745938563322e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8753093047507976, "epoch": 0.19454858648292633, "step": 480}, {"loss": 0.4005615234375, "grad_norm": 0.43917012214660645, "learning_rate": 9.293013314767284e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8670107159717776, "epoch": 0.19657513425879014, "step": 485}, {"loss": 0.3945556640625, "grad_norm": 0.398868203163147, "learning_rate": 9.276095715120727e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8710031307739684, "epoch": 0.19860168203465398, "step": 490}, {"loss": 0.378173828125, "grad_norm": 0.45519566535949707, "learning_rate": 9.25899386856325e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8744337476399202, "epoch": 0.20062822981051778, "step": 495}, {"loss": 0.3460693359375, "grad_norm": 0.4370631277561188, "learning_rate": 9.241708511973214e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8840001404214985, "epoch": 0.2026547775863816, "step": 500}, {"loss": 0.364599609375, "grad_norm": 0.3996580243110657, "learning_rate": 9.224240390135988e-06, "tIoU": 0.9998649597167969, "tRecall_30": 0.9998568534851074, "tRecall_50": 0.9998568534851074, "tRecall_70": 0.9998568534851074, "token_acc": 0.8795211009067047, "epoch": 0.20468132536224543, "step": 505}, {"loss": 0.3871826171875, "grad_norm": 0.48396146297454834, "learning_rate": 9.20659025571188e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8723862355459301, "epoch": 0.20670787313810923, "step": 510}, {"loss": 0.3907958984375, "grad_norm": 0.5639007091522217, "learning_rate": 9.188758869203682e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8702464822224618, "epoch": 0.20873442091397304, "step": 515}, {"loss": 0.3768310546875, "grad_norm": 0.46063733100891113, "learning_rate": 9.170746998923918e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8767145893566742, "epoch": 0.21076096868983688, "step": 520}, {"loss": 0.3778564453125, "grad_norm": 0.4143420457839966, "learning_rate": 9.152555420961729e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8748213349849921, "epoch": 0.21278751646570068, "step": 525}, {"loss": 0.406494140625, "grad_norm": 0.5685514211654663, "learning_rate": 9.13418491914944e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8642833706814105, "epoch": 0.2148140642415645, "step": 530}, {"loss": 0.3872314453125, "grad_norm": 0.5420283079147339, "learning_rate": 9.115636285028784e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8724378376307421, "epoch": 0.21684061201742832, "step": 535}, {"loss": 0.4113037109375, "grad_norm": 0.6400812864303589, "learning_rate": 9.096910317816796e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8705328273096042, "epoch": 0.21886715979329213, "step": 540}, {"loss": 0.40126953125, "grad_norm": 0.689870297908783, "learning_rate": 9.078007824371376e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8671837458878032, "epoch": 0.22089370756915594, "step": 545}, {"loss": 0.3592529296875, "grad_norm": 0.4439537227153778, "learning_rate": 9.058929619156527e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8815546864532758, "epoch": 0.22292025534501975, "step": 550}, {"loss": 0.37060546875, "grad_norm": 0.38134971261024475, "learning_rate": 9.039676524207255e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8778313570172185, "epoch": 0.22494680312088358, "step": 555}, {"loss": 0.40673828125, "grad_norm": 0.3596774637699127, "learning_rate": 9.020249369094159e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8652295664267323, "epoch": 0.2269733508967474, "step": 560}, {"loss": 0.3809326171875, "grad_norm": 0.6109125018119812, "learning_rate": 9.000648990887674e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8743921025504816, "epoch": 0.2289998986726112, "step": 565}, {"loss": 0.3836181640625, "grad_norm": 0.47461310029029846, "learning_rate": 8.980876234122017e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.874513549428124, "epoch": 0.23102644644847503, "step": 570}, {"loss": 0.347021484375, "grad_norm": 0.41040199995040894, "learning_rate": 8.960931950758792e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8863791737653685, "epoch": 0.23305299422433884, "step": 575}, {"loss": 0.3716064453125, "grad_norm": 0.4203317165374756, "learning_rate": 8.940817000150273e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8802946802696302, "epoch": 0.23507954200020265, "step": 580}, {"loss": 0.390380859375, "grad_norm": 0.446027934551239, "learning_rate": 8.920532249002394e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8720234062211272, "epoch": 0.23710608977606648, "step": 585}, {"loss": 0.4015625, "grad_norm": 0.9307211637496948, "learning_rate": 8.900078571337392e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8659462140807143, "epoch": 0.2391326375519303, "step": 590}, {"loss": 0.3928466796875, "grad_norm": 0.44321754574775696, "learning_rate": 8.879456848456145e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8695430999720835, "epoch": 0.2411591853277941, "step": 595}, {"loss": 0.370458984375, "grad_norm": 0.3784260153770447, "learning_rate": 8.85866796890021e-06, "tIoU": 0.999885082244873, "tRecall_30": 0.9998974800109863, "tRecall_50": 0.9998974800109863, "tRecall_70": 0.9998974800109863, "token_acc": 0.8785528921249093, "epoch": 0.24318573310365793, "step": 600}, {"loss": 0.3765380859375, "grad_norm": 0.45076823234558105, "learning_rate": 8.837712828413535e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8766975392882921, "epoch": 0.24521228087952174, "step": 605}, {"loss": 0.3655517578125, "grad_norm": 0.48391059041023254, "learning_rate": 8.81659232990385e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8825125394927084, "epoch": 0.24723882865538555, "step": 610}, {"loss": 0.375390625, "grad_norm": 0.9391803741455078, "learning_rate": 8.795307383403785e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8749314688760895, "epoch": 0.24926537643124938, "step": 615}, {"loss": 0.402587890625, "grad_norm": 0.5365148782730103, "learning_rate": 8.773858906031639e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8673238029359772, "epoch": 0.25129192420711316, "step": 620}, {"loss": 0.3646484375, "grad_norm": 0.4107930660247803, "learning_rate": 8.752247821951876e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8803534683936145, "epoch": 0.253318471982977, "step": 625}, {"loss": 0.403076171875, "grad_norm": 0.36408987641334534, "learning_rate": 8.730475062335298e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8675791889438111, "epoch": 0.25534501975884083, "step": 630}, {"loss": 0.37197265625, "grad_norm": 0.4031390845775604, "learning_rate": 8.70854156531893e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8779134815335501, "epoch": 0.25737156753470464, "step": 635}, {"loss": 0.3824951171875, "grad_norm": 0.486400842666626, "learning_rate": 8.68644827596559e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8733007408146467, "epoch": 0.25939811531056844, "step": 640}, {"loss": 0.3892578125, "grad_norm": 0.44566434621810913, "learning_rate": 8.664196146223172e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8704924857165266, "epoch": 0.26142466308643225, "step": 645}, {"loss": 0.382177734375, "grad_norm": 0.3780648112297058, "learning_rate": 8.641786134883631e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8741021634541154, "epoch": 0.26345121086229606, "step": 650}, {"loss": 0.3720947265625, "grad_norm": 0.44057774543762207, "learning_rate": 8.619219207541669e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8778030320781686, "epoch": 0.2654777586381599, "step": 655}, {"loss": 0.371533203125, "grad_norm": 0.4715423285961151, "learning_rate": 8.596496336553128e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8801489222275704, "epoch": 0.26750430641402373, "step": 660}, {"loss": 0.3737548828125, "grad_norm": 0.5181432962417603, "learning_rate": 8.573618500993095e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8751064228049731, "epoch": 0.26953085418988754, "step": 665}, {"loss": 0.365869140625, "grad_norm": 0.41301727294921875, "learning_rate": 8.550586686613715e-06, "tIoU": 0.9999523162841797, "tRecall_30": 0.9999526977539063, "tRecall_50": 0.9999526977539063, "tRecall_70": 0.9999526977539063, "token_acc": 0.878654232390975, "epoch": 0.27155740196575134, "step": 670}, {"loss": 0.3747802734375, "grad_norm": 0.473439484834671, "learning_rate": 8.527401885801723e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8762945655602113, "epoch": 0.27358394974161515, "step": 675}, {"loss": 0.383251953125, "grad_norm": 0.425710529088974, "learning_rate": 8.504065097535674e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8726958420525884, "epoch": 0.27561049751747896, "step": 680}, {"loss": 0.404248046875, "grad_norm": 0.4751035273075104, "learning_rate": 8.48057732734291e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8653982002918089, "epoch": 0.27763704529334277, "step": 685}, {"loss": 0.381005859375, "grad_norm": 0.44467607140541077, "learning_rate": 8.456939587256225e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8726581191337954, "epoch": 0.27966359306920663, "step": 690}, {"loss": 0.39541015625, "grad_norm": 0.48777613043785095, "learning_rate": 8.433152895770269e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8681833845913769, "epoch": 0.28169014084507044, "step": 695}, {"loss": 0.3770263671875, "grad_norm": 0.4859802722930908, "learning_rate": 8.409218277797648e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8738482211415408, "epoch": 0.28371668862093424, "step": 700}, {"loss": 0.3587890625, "grad_norm": 0.3302393853664398, "learning_rate": 8.385136764624782e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8811119122569451, "epoch": 0.28574323639679805, "step": 705}, {"loss": 0.3962158203125, "grad_norm": 0.39953362941741943, "learning_rate": 8.360909393867455e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8683007948043105, "epoch": 0.28776978417266186, "step": 710}, {"loss": 0.37705078125, "grad_norm": 0.2610228955745697, "learning_rate": 8.336537209426115e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8759010268027891, "epoch": 0.28979633194852567, "step": 715}, {"loss": 0.360986328125, "grad_norm": 0.5030771493911743, "learning_rate": 8.312021261440883e-06, "tIoU": 0.9996023178100586, "tRecall_30": 0.9997917175292969, "tRecall_50": 0.9997917175292969, "tRecall_70": 0.9997917175292969, "token_acc": 0.8792668163208196, "epoch": 0.29182287972438953, "step": 720}, {"loss": 0.3444091796875, "grad_norm": 0.4199218153953552, "learning_rate": 8.287362606246325e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8850844306174159, "epoch": 0.29384942750025334, "step": 725}, {"loss": 0.380419921875, "grad_norm": 0.43417927622795105, "learning_rate": 8.262562306325913e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8734411766591651, "epoch": 0.29587597527611714, "step": 730}, {"loss": 0.356982421875, "grad_norm": 0.40655168890953064, "learning_rate": 8.237621430266267e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8851337543171275, "epoch": 0.29790252305198095, "step": 735}, {"loss": 0.351123046875, "grad_norm": 0.35124924778938293, "learning_rate": 8.212541052711098e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8840490141313778, "epoch": 0.29992907082784476, "step": 740}, {"loss": 0.4010498046875, "grad_norm": 0.37331146001815796, "learning_rate": 8.187322254314907e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8676304091337044, "epoch": 0.30195561860370856, "step": 745}, {"loss": 0.3976318359375, "grad_norm": 0.536299467086792, "learning_rate": 8.16196612169643e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8687691912291855, "epoch": 0.30398216637957237, "step": 750}, {"loss": 0.37802734375, "grad_norm": 0.7118610143661499, "learning_rate": 8.136473747391802e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8742820838627701, "epoch": 0.30600871415543623, "step": 755}, {"loss": 0.3813232421875, "grad_norm": 0.4947682023048401, "learning_rate": 8.110846229807497e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8758169106041493, "epoch": 0.30803526193130004, "step": 760}, {"loss": 0.384326171875, "grad_norm": 0.36296963691711426, "learning_rate": 8.085084673173e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8708688664817646, "epoch": 0.31006180970716385, "step": 765}, {"loss": 0.3591796875, "grad_norm": 0.4178961217403412, "learning_rate": 8.059190187493216e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8794477029858403, "epoch": 0.31208835748302766, "step": 770}, {"loss": 0.3423828125, "grad_norm": 0.47259947657585144, "learning_rate": 8.033163888500653e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.884736740554099, "epoch": 0.31411490525889146, "step": 775}, {"loss": 0.3815673828125, "grad_norm": 0.36307206749916077, "learning_rate": 8.007006897607346e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8737486629549381, "epoch": 0.31614145303475527, "step": 780}, {"loss": 0.370849609375, "grad_norm": 0.47526979446411133, "learning_rate": 7.980720341856532e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8770150182461645, "epoch": 0.31816800081061913, "step": 785}, {"loss": 0.3899658203125, "grad_norm": 0.43980148434638977, "learning_rate": 7.9543053538741e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8706937049796429, "epoch": 0.32019454858648294, "step": 790}, {"loss": 0.336474609375, "grad_norm": 0.4804976284503937, "learning_rate": 7.927763071819779e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8857131373911246, "epoch": 0.32222109636234675, "step": 795}, {"loss": 0.4080322265625, "grad_norm": 0.5127283334732056, "learning_rate": 7.901094639338098e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8650576159254072, "epoch": 0.32424764413821056, "step": 800}, {"loss": 0.3904296875, "grad_norm": 0.5694597363471985, "learning_rate": 7.874301205509112e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8694054648160928, "epoch": 0.32627419191407436, "step": 805}, {"loss": 0.3904541015625, "grad_norm": 0.43124932050704956, "learning_rate": 7.847383924798892e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.871555706607253, "epoch": 0.32830073968993817, "step": 810}, {"loss": 0.390966796875, "grad_norm": 0.4303266406059265, "learning_rate": 7.82034395700978e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8692795917031997, "epoch": 0.33032728746580203, "step": 815}, {"loss": 0.3556884765625, "grad_norm": 0.3481966257095337, "learning_rate": 7.793182467230415e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8809747012649367, "epoch": 0.33235383524166584, "step": 820}, {"loss": 0.3641357421875, "grad_norm": 0.5233892202377319, "learning_rate": 7.765900625785532e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8789725840008653, "epoch": 0.33438038301752965, "step": 825}, {"loss": 0.3589599609375, "grad_norm": 0.4372534453868866, "learning_rate": 7.738499608185536e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8846019057790832, "epoch": 0.33640693079339345, "step": 830}, {"loss": 0.376806640625, "grad_norm": 0.42524704337120056, "learning_rate": 7.710980595075855e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8755772871051479, "epoch": 0.33843347856925726, "step": 835}, {"loss": 0.33037109375, "grad_norm": 0.28774598240852356, "learning_rate": 7.683344772186063e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.892224320921029, "epoch": 0.34046002634512107, "step": 840}, {"loss": 0.3840087890625, "grad_norm": 0.4768100380897522, "learning_rate": 7.655593330278798e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8781859667120646, "epoch": 0.3424865741209849, "step": 845}, {"loss": 0.360986328125, "grad_norm": 0.37228870391845703, "learning_rate": 7.627727465098442e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8804808702639224, "epoch": 0.34451312189684874, "step": 850}, {"loss": 0.383642578125, "grad_norm": 0.42258158326148987, "learning_rate": 7.599748377319611e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8718607702866956, "epoch": 0.34653966967271255, "step": 855}, {"loss": 0.403564453125, "grad_norm": 0.5075936913490295, "learning_rate": 7.5716572724954184e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8665800600833172, "epoch": 0.34856621744857635, "step": 860}, {"loss": 0.3584716796875, "grad_norm": 0.3515145182609558, "learning_rate": 7.543455361005526e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8810606243803352, "epoch": 0.35059276522444016, "step": 865}, {"loss": 0.3622802734375, "grad_norm": 0.4498981833457947, "learning_rate": 7.5151438580039945e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8861151058022481, "epoch": 0.35261931300030397, "step": 870}, {"loss": 0.3845703125, "grad_norm": 0.4505605697631836, "learning_rate": 7.486723983366925e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8717318269256821, "epoch": 0.3546458607761678, "step": 875}, {"loss": 0.380859375, "grad_norm": 0.6111229062080383, "learning_rate": 7.4581969616399e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8744601628693373, "epoch": 0.35667240855203164, "step": 880}, {"loss": 0.4007568359375, "grad_norm": 0.5495724678039551, "learning_rate": 7.4295640219852135e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8689283873573946, "epoch": 0.35869895632789545, "step": 885}, {"loss": 0.3607421875, "grad_norm": 0.3697036802768707, "learning_rate": 7.40082639812892e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8797935193453291, "epoch": 0.36072550410375925, "step": 890}, {"loss": 0.3994873046875, "grad_norm": 0.5050045847892761, "learning_rate": 7.371985328307664e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8684937988743339, "epoch": 0.36275205187962306, "step": 895}, {"loss": 0.381591796875, "grad_norm": 0.42693156003952026, "learning_rate": 7.343042055215342e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8741249800223749, "epoch": 0.36477859965548687, "step": 900}, {"loss": 0.37421875, "grad_norm": 0.4217187762260437, "learning_rate": 7.313997825949536e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8787105693100414, "epoch": 0.3668051474313507, "step": 905}, {"loss": 0.3683837890625, "grad_norm": 0.41110745072364807, "learning_rate": 7.284853891957806e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8784763482289679, "epoch": 0.3688316952072145, "step": 910}, {"loss": 0.3561767578125, "grad_norm": 0.43751952052116394, "learning_rate": 7.2556115089837435e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8809951857527238, "epoch": 0.37085824298307835, "step": 915}, {"loss": 0.396240234375, "grad_norm": 0.3949474096298218, "learning_rate": 7.226271937012883e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8673576077580574, "epoch": 0.37288479075894215, "step": 920}, {"loss": 0.3632568359375, "grad_norm": 0.4348854124546051, "learning_rate": 7.1968364402183985e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8787858087673649, "epoch": 0.37491133853480596, "step": 925}, {"loss": 0.3761962890625, "grad_norm": 0.45749151706695557, "learning_rate": 7.167306286906645e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8756596442196003, "epoch": 0.37693788631066977, "step": 930}, {"loss": 0.394189453125, "grad_norm": 0.4483352601528168, "learning_rate": 7.137682749462496e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8667820435842193, "epoch": 0.3789644340865336, "step": 935}, {"loss": 0.3496337890625, "grad_norm": 0.5248214602470398, "learning_rate": 7.107967104294538e-06, "tIoU": 0.99993896484375, "tRecall_30": 0.9999264717102051, "tRecall_50": 0.9999264717102051, "tRecall_70": 0.9999264717102051, "token_acc": 0.8841316967787876, "epoch": 0.3809909818623974, "step": 940}, {"loss": 0.37080078125, "grad_norm": 0.39933234453201294, "learning_rate": 7.078160631780054e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8746478070647529, "epoch": 0.38301752963826124, "step": 945}, {"loss": 0.365966796875, "grad_norm": 0.49557754397392273, "learning_rate": 7.048264616209869e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8790204772034137, "epoch": 0.38504407741412505, "step": 950}, {"loss": 0.3338623046875, "grad_norm": 0.4162745773792267, "learning_rate": 7.018280345733004e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.888543501924566, "epoch": 0.38707062518998886, "step": 955}, {"loss": 0.338916015625, "grad_norm": 0.3654502332210541, "learning_rate": 6.988209112301181e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8869780634043012, "epoch": 0.38909717296585267, "step": 960}, {"loss": 0.3736083984375, "grad_norm": 0.3963611125946045, "learning_rate": 6.958052211613147e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8755711663818746, "epoch": 0.3911237207417165, "step": 965}, {"loss": 0.3814208984375, "grad_norm": 0.5171037912368774, "learning_rate": 6.9278109430588505e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8742617061438227, "epoch": 0.3931502685175803, "step": 970}, {"loss": 0.365283203125, "grad_norm": 0.5404708385467529, "learning_rate": 6.897486609663455e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.882941194328911, "epoch": 0.39517681629344414, "step": 975}, {"loss": 0.39111328125, "grad_norm": 0.527686357498169, "learning_rate": 6.867080518031194e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8707319961822223, "epoch": 0.39720336406930795, "step": 980}, {"loss": 0.3810302734375, "grad_norm": 0.5123360753059387, "learning_rate": 6.836593978289064e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8738838872216664, "epoch": 0.39922991184517176, "step": 985}, {"loss": 0.3386962890625, "grad_norm": 0.37030547857284546, "learning_rate": 6.80602830403039e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8885187714839158, "epoch": 0.40125645962103557, "step": 990}, {"loss": 0.3621337890625, "grad_norm": 0.5129815936088562, "learning_rate": 6.775384812258212e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8785935775476954, "epoch": 0.4032830073968994, "step": 995}, {"loss": 0.3682861328125, "grad_norm": 0.48550745844841003, "learning_rate": 6.744664823328545e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8782230637497287, "epoch": 0.4053095551727632, "step": 1000}, {"loss": 0.3373046875, "grad_norm": 0.360082745552063, "learning_rate": 6.713869660893486e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8856287287482901, "epoch": 0.407336102948627, "step": 1005}, {"loss": 0.3563232421875, "grad_norm": 0.37603187561035156, "learning_rate": 6.683000651844186e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.880387902345518, "epoch": 0.40936265072449085, "step": 1010}, {"loss": 0.3497314453125, "grad_norm": 0.3709280490875244, "learning_rate": 6.652059126253665e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.886997627845445, "epoch": 0.41138919850035466, "step": 1015}, {"loss": 0.381396484375, "grad_norm": 0.4052375555038452, "learning_rate": 6.621046417319518e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8730041806398253, "epoch": 0.41341574627621847, "step": 1020}, {"loss": 0.3515869140625, "grad_norm": 0.39146503806114197, "learning_rate": 6.589963861306456e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8837771654074887, "epoch": 0.4154422940520823, "step": 1025}, {"loss": 0.3480712890625, "grad_norm": 0.40089619159698486, "learning_rate": 6.558812797488745e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8839427066060109, "epoch": 0.4174688418279461, "step": 1030}, {"loss": 0.3528564453125, "grad_norm": 0.4794605076313019, "learning_rate": 6.527594568092482e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8843052146682948, "epoch": 0.4194953896038099, "step": 1035}, {"loss": 0.33486328125, "grad_norm": 0.5362907648086548, "learning_rate": 6.4963105182377784e-06, "tIoU": 0.9999542236328125, "tRecall_30": 0.9999536514282227, "tRecall_50": 0.9999536514282227, "tRecall_70": 0.9999536514282227, "token_acc": 0.8922635725746582, "epoch": 0.42152193737967375, "step": 1040}, {"loss": 0.37939453125, "grad_norm": 0.35571110248565674, "learning_rate": 6.464961995880788e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.877855904146303, "epoch": 0.42354848515553756, "step": 1045}, {"loss": 0.3494384765625, "grad_norm": 0.36828193068504333, "learning_rate": 6.433550351755639e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8832511866683361, "epoch": 0.42557503293140136, "step": 1050}, {"loss": 0.3216064453125, "grad_norm": 0.35322102904319763, "learning_rate": 6.402076939316223e-06, "tIoU": 0.99375, "tRecall_30": 0.99375, "tRecall_50": 0.99375, "tRecall_70": 0.99375, "token_acc": 0.8966078344163241, "epoch": 0.42760158070726517, "step": 1055}, {"loss": 0.3531982421875, "grad_norm": 0.5031628608703613, "learning_rate": 6.370543114677887e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8816162500961772, "epoch": 0.429628128483129, "step": 1060}, {"loss": 0.3392578125, "grad_norm": 0.409626305103302, "learning_rate": 6.338950236558996e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8864688797464766, "epoch": 0.4316546762589928, "step": 1065}, {"loss": 0.3597412109375, "grad_norm": 0.41023844480514526, "learning_rate": 6.307299666222391e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8799794250442933, "epoch": 0.43368122403485665, "step": 1070}, {"loss": 0.3640380859375, "grad_norm": 0.4652061462402344, "learning_rate": 6.275592767416734e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8776116953453434, "epoch": 0.43570777181072046, "step": 1075}, {"loss": 0.36923828125, "grad_norm": 0.4822935461997986, "learning_rate": 6.243830906317749e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8774388126503158, "epoch": 0.43773431958658426, "step": 1080}, {"loss": 0.3608154296875, "grad_norm": 0.4883968234062195, "learning_rate": 6.212015451469357e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8790962323901034, "epoch": 0.43976086736244807, "step": 1085}, {"loss": 0.3548828125, "grad_norm": 0.43007761240005493, "learning_rate": 6.180147773724706e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8837881721101605, "epoch": 0.4417874151383119, "step": 1090}, {"loss": 0.3470703125, "grad_norm": 0.4888928234577179, "learning_rate": 6.148229246187107e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88543772516882, "epoch": 0.4438139629141757, "step": 1095}, {"loss": 0.396435546875, "grad_norm": 0.4666309058666229, "learning_rate": 6.116261244150868e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8679060399031002, "epoch": 0.4458405106900395, "step": 1100}, {"loss": 0.353662109375, "grad_norm": 0.42913663387298584, "learning_rate": 6.084245145042039e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8826747113821377, "epoch": 0.44786705846590336, "step": 1105}, {"loss": 0.33896484375, "grad_norm": 0.482410728931427, "learning_rate": 6.052182328359056e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8857838859345668, "epoch": 0.44989360624176716, "step": 1110}, {"loss": 0.3577880859375, "grad_norm": 0.4276754558086395, "learning_rate": 6.0200741756133105e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8800624976841274, "epoch": 0.45192015401763097, "step": 1115}, {"loss": 0.32001953125, "grad_norm": 0.4833751320838928, "learning_rate": 5.987922070269612e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8940394473889999, "epoch": 0.4539467017934948, "step": 1120}, {"loss": 0.3525634765625, "grad_norm": 0.4228677451610565, "learning_rate": 5.95572739768659e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8830779480346436, "epoch": 0.4559732495693586, "step": 1125}, {"loss": 0.3637939453125, "grad_norm": 0.4276502728462219, "learning_rate": 5.923491545056992e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8791735003721488, "epoch": 0.4579997973452224, "step": 1130}, {"loss": 0.3595947265625, "grad_norm": 0.46918392181396484, "learning_rate": 5.891215901347922e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8808175171468626, "epoch": 0.46002634512108626, "step": 1135}, {"loss": 0.3795166015625, "grad_norm": 0.43709152936935425, "learning_rate": 5.85890185724098e-06, "tIoU": 0.99375, "tRecall_30": 0.99375, "tRecall_50": 0.99375, "tRecall_70": 0.99375, "token_acc": 0.872928022166402, "epoch": 0.46205289289695006, "step": 1140}, {"loss": 0.3873046875, "grad_norm": 0.4206150472164154, "learning_rate": 5.826550805072358e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8712319938126389, "epoch": 0.46407944067281387, "step": 1145}, {"loss": 0.337109375, "grad_norm": 0.3167666494846344, "learning_rate": 5.794164138772832e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8865754147337275, "epoch": 0.4661059884486777, "step": 1150}, {"loss": 0.36435546875, "grad_norm": 0.4532758891582489, "learning_rate": 5.76174325380771e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8779640518250711, "epoch": 0.4681325362245415, "step": 1155}, {"loss": 0.389697265625, "grad_norm": 0.4313461184501648, "learning_rate": 5.729289547116701e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8700403579575364, "epoch": 0.4701590840004053, "step": 1160}, {"loss": 0.3765380859375, "grad_norm": 0.4814442992210388, "learning_rate": 5.696804417053724e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8757619796016335, "epoch": 0.4721856317762691, "step": 1165}, {"loss": 0.3399658203125, "grad_norm": 0.3513696491718292, "learning_rate": 5.664289263326659e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8893098409021976, "epoch": 0.47421217955213296, "step": 1170}, {"loss": 0.3510986328125, "grad_norm": 0.40614351630210876, "learning_rate": 5.631745486937035e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8833963519767688, "epoch": 0.47623872732799677, "step": 1175}, {"loss": 0.3646240234375, "grad_norm": 0.5500504374504089, "learning_rate": 5.599174490119661e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8787946444647706, "epoch": 0.4782652751038606, "step": 1180}, {"loss": 0.3532470703125, "grad_norm": 0.5975911021232605, "learning_rate": 5.566577676282216e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8857554384480847, "epoch": 0.4802918228797244, "step": 1185}, {"loss": 0.32177734375, "grad_norm": 0.3443179130554199, "learning_rate": 5.533956449944771e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8915946582875098, "epoch": 0.4823183706555882, "step": 1190}, {"loss": 0.346630859375, "grad_norm": 0.3399013578891754, "learning_rate": 5.501312216679271e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8862332459567733, "epoch": 0.484344918431452, "step": 1195}, {"loss": 0.3672119140625, "grad_norm": 0.4984676241874695, "learning_rate": 5.468646383048978e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8770362908179562, "epoch": 0.48637146620731586, "step": 1200}, {"loss": 0.3676025390625, "grad_norm": 0.3607572913169861, "learning_rate": 5.435960356547866e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.878624072395896, "epoch": 0.48839801398317967, "step": 1205}, {"loss": 0.3791259765625, "grad_norm": 0.4026191234588623, "learning_rate": 5.403255545539967e-06, "tIoU": 0.9997499465942383, "tRecall_30": 0.9995536804199219, "tRecall_50": 0.9995536804199219, "tRecall_70": 0.9995536804199219, "token_acc": 0.874137520818463, "epoch": 0.4904245617590435, "step": 1210}, {"loss": 0.3332763671875, "grad_norm": 0.4158448576927185, "learning_rate": 5.370533359198697e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8925485768613702, "epoch": 0.4924511095349073, "step": 1215}, {"loss": 0.3458740234375, "grad_norm": 0.4042837917804718, "learning_rate": 5.337795207446134e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8848929158802503, "epoch": 0.4944776573107711, "step": 1220}, {"loss": 0.334619140625, "grad_norm": 0.38991838693618774, "learning_rate": 5.305042500892265e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8871413548783782, "epoch": 0.4965042050866349, "step": 1225}, {"loss": 0.383642578125, "grad_norm": 0.38406676054000854, "learning_rate": 5.272276650774211e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8712314225053078, "epoch": 0.49853075286249876, "step": 1230}, {"loss": 0.3812255859375, "grad_norm": 0.4755161702632904, "learning_rate": 5.23949906889542e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8735439060691251, "epoch": 0.5005573006383626, "step": 1235}, {"loss": 0.3880126953125, "grad_norm": 0.48450133204460144, "learning_rate": 5.20671116756483e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8706281504459092, "epoch": 0.5025838484142263, "step": 1240}, {"loss": 0.3583251953125, "grad_norm": 0.4431735873222351, "learning_rate": 5.173914359536026e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8804128704576395, "epoch": 0.5046103961900902, "step": 1245}, {"loss": 0.353564453125, "grad_norm": 0.4925408959388733, "learning_rate": 5.141110057946357e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.881574081411628, "epoch": 0.506636943965954, "step": 1250}, {"loss": 0.36533203125, "grad_norm": 0.39856627583503723, "learning_rate": 5.1082996762560495e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8775912245354824, "epoch": 0.5086634917418178, "step": 1255}, {"loss": 0.35595703125, "grad_norm": 0.41317877173423767, "learning_rate": 5.075484628187315e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8816695158533046, "epoch": 0.5106900395176817, "step": 1260}, {"loss": 0.3582763671875, "grad_norm": 0.4024195969104767, "learning_rate": 5.042666327663421e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8803793129712849, "epoch": 0.5127165872935454, "step": 1265}, {"loss": 0.3602783203125, "grad_norm": 0.513283371925354, "learning_rate": 5.009846188747782e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8798386381166255, "epoch": 0.5147431350694093, "step": 1270}, {"loss": 0.376123046875, "grad_norm": 0.44748246669769287, "learning_rate": 4.977025625583017e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8758046600018736, "epoch": 0.516769682845273, "step": 1275}, {"loss": 0.35078125, "grad_norm": 0.3967839777469635, "learning_rate": 4.944206052330032e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8834837609267288, "epoch": 0.5187962306211369, "step": 1280}, {"loss": 0.3705810546875, "grad_norm": 0.3833644986152649, "learning_rate": 4.911388883107076e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8753450905744576, "epoch": 0.5208227783970008, "step": 1285}, {"loss": 0.3374267578125, "grad_norm": 0.3436470031738281, "learning_rate": 4.878575531928815e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8864620406000278, "epoch": 0.5228493261728645, "step": 1290}, {"loss": 0.3248046875, "grad_norm": 0.6544244885444641, "learning_rate": 4.845767412645405e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.891281923817809, "epoch": 0.5248758739487284, "step": 1295}, {"loss": 0.3455322265625, "grad_norm": 0.378275066614151, "learning_rate": 4.81296593888157e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8847039987503905, "epoch": 0.5269024217245921, "step": 1300}, {"loss": 0.3330810546875, "grad_norm": 0.4009753167629242, "learning_rate": 4.780172523975697e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8884986977137642, "epoch": 0.528928969500456, "step": 1305}, {"loss": 0.3642333984375, "grad_norm": 0.46430066227912903, "learning_rate": 4.747388580918934e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8783965865708511, "epoch": 0.5309555172763198, "step": 1310}, {"loss": 0.3664794921875, "grad_norm": 0.49786871671676636, "learning_rate": 4.71461552229431e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8775798506769107, "epoch": 0.5329820650521836, "step": 1315}, {"loss": 0.3385986328125, "grad_norm": 0.46610528230667114, "learning_rate": 4.68185476021587e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8856598940033139, "epoch": 0.5350086128280475, "step": 1320}, {"loss": 0.3858642578125, "grad_norm": 0.45692896842956543, "learning_rate": 4.649107706267823e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8724768936577074, "epoch": 0.5370351606039112, "step": 1325}, {"loss": 0.37216796875, "grad_norm": 0.5358167886734009, "learning_rate": 4.616375771443743e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8778596081550957, "epoch": 0.5390617083797751, "step": 1330}, {"loss": 0.3567626953125, "grad_norm": 0.4610866904258728, "learning_rate": 4.583660366085735e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8821897267269218, "epoch": 0.5410882561556388, "step": 1335}, {"loss": 0.3708984375, "grad_norm": 0.4700048863887787, "learning_rate": 4.5509628998237085e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8747563915390623, "epoch": 0.5431148039315027, "step": 1340}, {"loss": 0.371826171875, "grad_norm": 0.4478241503238678, "learning_rate": 4.518284781514603e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8773185057988198, "epoch": 0.5451413517073666, "step": 1345}, {"loss": 0.3501953125, "grad_norm": 0.4090299904346466, "learning_rate": 4.485627419181714e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8831489211736755, "epoch": 0.5471678994832303, "step": 1350}, {"loss": 0.34912109375, "grad_norm": 0.4647216200828552, "learning_rate": 4.452992219953999e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8830212417158579, "epoch": 0.5491944472590942, "step": 1355}, {"loss": 0.350537109375, "grad_norm": 0.5491527318954468, "learning_rate": 4.42038059000547e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8814508260152754, "epoch": 0.5512209950349579, "step": 1360}, {"loss": 0.3564453125, "grad_norm": 0.3985271751880646, "learning_rate": 4.387793934494583e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8810057091206359, "epoch": 0.5532475428108218, "step": 1365}, {"loss": 0.338720703125, "grad_norm": 0.38414013385772705, "learning_rate": 4.355233657503717e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8854592141406903, "epoch": 0.5552740905866855, "step": 1370}, {"loss": 0.373388671875, "grad_norm": 0.467517614364624, "learning_rate": 4.322701161978648e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8745456194576039, "epoch": 0.5573006383625494, "step": 1375}, {"loss": 0.3896728515625, "grad_norm": 0.41305845975875854, "learning_rate": 4.29019784966813e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8690092846025869, "epoch": 0.5593271861384133, "step": 1380}, {"loss": 0.3476318359375, "grad_norm": 0.45166894793510437, "learning_rate": 4.257725121063464e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8838776329382966, "epoch": 0.561353733914277, "step": 1385}, {"loss": 0.3684814453125, "grad_norm": 0.3921132981777191, "learning_rate": 4.225284375338187e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8769061119557345, "epoch": 0.5633802816901409, "step": 1390}, {"loss": 0.3556640625, "grad_norm": 0.6856555938720703, "learning_rate": 4.192877010287755e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8795452874700099, "epoch": 0.5654068294660046, "step": 1395}, {"loss": 0.3482666015625, "grad_norm": 0.4415869414806366, "learning_rate": 4.160504422269343e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8845478325859492, "epoch": 0.5674333772418685, "step": 1400}, {"loss": 0.359033203125, "grad_norm": 0.44592997431755066, "learning_rate": 4.128168006141651e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8789114899636594, "epoch": 0.5694599250177322, "step": 1405}, {"loss": 0.3491943359375, "grad_norm": 0.420045405626297, "learning_rate": 4.09586915520483e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.882240717334796, "epoch": 0.5714864727935961, "step": 1410}, {"loss": 0.36396484375, "grad_norm": 0.348161518573761, "learning_rate": 4.063609261140425e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8796527000915285, "epoch": 0.57351302056946, "step": 1415}, {"loss": 0.3662109375, "grad_norm": 0.5775684118270874, "learning_rate": 4.031389713951432e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8783637725837949, "epoch": 0.5755395683453237, "step": 1420}, {"loss": 0.369384765625, "grad_norm": 0.39185187220573425, "learning_rate": 3.999211901902382e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8774693489029958, "epoch": 0.5775661161211876, "step": 1425}, {"loss": 0.349609375, "grad_norm": 0.49676889181137085, "learning_rate": 3.967077211459547e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8816592219550906, "epoch": 0.5795926638970513, "step": 1430}, {"loss": 0.360302734375, "grad_norm": 0.42753615975379944, "learning_rate": 3.934987027231185e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8803608665251746, "epoch": 0.5816192116729152, "step": 1435}, {"loss": 0.3525634765625, "grad_norm": 0.41853469610214233, "learning_rate": 3.90294273190789e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8811438899178998, "epoch": 0.5836457594487791, "step": 1440}, {"loss": 0.3712646484375, "grad_norm": 0.3808477818965912, "learning_rate": 3.870945706203008e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8750312781503353, "epoch": 0.5856723072246428, "step": 1445}, {"loss": 0.3540283203125, "grad_norm": 0.3081310987472534, "learning_rate": 3.8389973287931486e-06, "tIoU": 0.9997791290283203, "tRecall_30": 0.9997023582458496, "tRecall_50": 0.9997023582458496, "tRecall_70": 0.9997023582458496, "token_acc": 0.8833477455219271, "epoch": 0.5876988550005067, "step": 1450}, {"loss": 0.3547607421875, "grad_norm": 0.3728139102458954, "learning_rate": 3.8070989762587805e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8799670478774957, "epoch": 0.5897254027763704, "step": 1455}, {"loss": 0.354736328125, "grad_norm": 0.5454410910606384, "learning_rate": 3.7752520230249216e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8818026764628706, "epoch": 0.5917519505522343, "step": 1460}, {"loss": 0.3484619140625, "grad_norm": 0.3359200656414032, "learning_rate": 3.743457841301909e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.881686928154548, "epoch": 0.593778498328098, "step": 1465}, {"loss": 0.36337890625, "grad_norm": 0.4866096079349518, "learning_rate": 3.7117178010262843e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8767980029235277, "epoch": 0.5958050461039619, "step": 1470}, {"loss": 0.3525390625, "grad_norm": 0.42000898718833923, "learning_rate": 3.680033269801759e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8828758953386568, "epoch": 0.5978315938798258, "step": 1475}, {"loss": 0.317041015625, "grad_norm": 0.4449663460254669, "learning_rate": 3.648405612840291e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8941719450487374, "epoch": 0.5998581416556895, "step": 1480}, {"loss": 0.3648193359375, "grad_norm": 0.4433095157146454, "learning_rate": 3.6168361929032584e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8779625300451088, "epoch": 0.6018846894315534, "step": 1485}, {"loss": 0.336474609375, "grad_norm": 0.31316468119621277, "learning_rate": 3.5853263702427465e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8874035573171796, "epoch": 0.6039112372074171, "step": 1490}, {"loss": 0.34521484375, "grad_norm": 0.3436121940612793, "learning_rate": 3.5538775025429285e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8830729953564668, "epoch": 0.605937784983281, "step": 1495}, {"loss": 0.3583984375, "grad_norm": 0.3756560981273651, "learning_rate": 3.522490944861578e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8821701099263564, "epoch": 0.6079643327591447, "step": 1500}, {"loss": 0.3664794921875, "grad_norm": 0.43999379873275757, "learning_rate": 3.4911680495716703e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8774288485334939, "epoch": 0.6099908805350086, "step": 1505}, {"loss": 0.3483642578125, "grad_norm": 0.48006972670555115, "learning_rate": 3.459910166303123e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8825045995932991, "epoch": 0.6120174283108725, "step": 1510}, {"loss": 0.3367919921875, "grad_norm": 0.37062063813209534, "learning_rate": 3.4287186418846353e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8869997197058701, "epoch": 0.6140439760867362, "step": 1515}, {"loss": 0.3341064453125, "grad_norm": 0.45247992873191833, "learning_rate": 3.397594820285661e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8887578872530142, "epoch": 0.6160705238626001, "step": 1520}, {"loss": 0.3563232421875, "grad_norm": 0.3768942654132843, "learning_rate": 3.366540042558497e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8823201159673136, "epoch": 0.6180970716384638, "step": 1525}, {"loss": 0.350146484375, "grad_norm": 0.4086240828037262, "learning_rate": 3.3355556467805063e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8824231340428154, "epoch": 0.6201236194143277, "step": 1530}, {"loss": 0.3452880859375, "grad_norm": 0.39199697971343994, "learning_rate": 3.304642967996452e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8839277431803143, "epoch": 0.6221501671901916, "step": 1535}, {"loss": 0.3403564453125, "grad_norm": 0.44045987725257874, "learning_rate": 3.2738033381609887e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.885882583705324, "epoch": 0.6241767149660553, "step": 1540}, {"loss": 0.3364990234375, "grad_norm": 0.3382072448730469, "learning_rate": 3.243038086081256e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8862242631575911, "epoch": 0.6262032627419192, "step": 1545}, {"loss": 0.3208984375, "grad_norm": 0.33277419209480286, "learning_rate": 3.2123485373596387e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8928338244895798, "epoch": 0.6282298105177829, "step": 1550}, {"loss": 0.3424072265625, "grad_norm": 0.5623494982719421, "learning_rate": 3.181736014336633e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8854034061041338, "epoch": 0.6302563582936468, "step": 1555}, {"loss": 0.3233154296875, "grad_norm": 0.40705209970474243, "learning_rate": 3.151201836033887e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8919693124471282, "epoch": 0.6322829060695105, "step": 1560}, {"loss": 0.32705078125, "grad_norm": 0.3886021077632904, "learning_rate": 3.1207473180973537e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8905853670549413, "epoch": 0.6343094538453744, "step": 1565}, {"loss": 0.340234375, "grad_norm": 0.45216310024261475, "learning_rate": 3.0903737727406103e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8851567758256551, "epoch": 0.6363360016212383, "step": 1570}, {"loss": 0.345654296875, "grad_norm": 0.42403921484947205, "learning_rate": 3.0600825086883157e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8841203080129101, "epoch": 0.638362549397102, "step": 1575}, {"loss": 0.3524658203125, "grad_norm": 0.43884193897247314, "learning_rate": 3.029874831119821e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8821244395373574, "epoch": 0.6403890971729659, "step": 1580}, {"loss": 0.3519287109375, "grad_norm": 0.38225412368774414, "learning_rate": 2.9997520416129317e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.881298560950009, "epoch": 0.6424156449488296, "step": 1585}, {"loss": 0.3475830078125, "grad_norm": 0.4440082311630249, "learning_rate": 2.9697154380878267e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8843408659299762, "epoch": 0.6444421927246935, "step": 1590}, {"loss": 0.3405517578125, "grad_norm": 0.38897669315338135, "learning_rate": 2.9397663147511333e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8871316466216582, "epoch": 0.6464687405005572, "step": 1595}, {"loss": 0.3591064453125, "grad_norm": 0.4603720009326935, "learning_rate": 2.9099059620401647e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8784650512461998, "epoch": 0.6484952882764211, "step": 1600}, {"loss": 0.3609130859375, "grad_norm": 0.4842081665992737, "learning_rate": 2.88013566656731e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8791378525786873, "epoch": 0.650521836052285, "step": 1605}, {"loss": 0.3239501953125, "grad_norm": 0.3952023983001709, "learning_rate": 2.8504567110646166e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8904520420960794, "epoch": 0.6525483838281487, "step": 1610}, {"loss": 0.3415283203125, "grad_norm": 0.4561847746372223, "learning_rate": 2.8208703743284928e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8860618126198743, "epoch": 0.6545749316040126, "step": 1615}, {"loss": 0.3582275390625, "grad_norm": 0.47599977254867554, "learning_rate": 2.7913779311646337e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8794347726568916, "epoch": 0.6566014793798763, "step": 1620}, {"loss": 0.320068359375, "grad_norm": 0.4711098372936249, "learning_rate": 2.761980652333073e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8925188697393882, "epoch": 0.6586280271557402, "step": 1625}, {"loss": 0.324853515625, "grad_norm": 0.47064128518104553, "learning_rate": 2.73267980449344e-06, "tIoU": 0.9998732566833496, "tRecall_30": 0.9995833396911621, "tRecall_50": 0.9995833396911621, "tRecall_70": 0.9995833396911621, "token_acc": 0.8897695675796606, "epoch": 0.6606545749316041, "step": 1630}, {"loss": 0.3352783203125, "grad_norm": 0.47430020570755005, "learning_rate": 2.70347665015038e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8870509525156498, "epoch": 0.6626811227074678, "step": 1635}, {"loss": 0.3516845703125, "grad_norm": 0.42994794249534607, "learning_rate": 2.6743724475991584e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.881572338841925, "epoch": 0.6647076704833317, "step": 1640}, {"loss": 0.3171630859375, "grad_norm": 0.37945184111595154, "learning_rate": 2.6453684508714307e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8962419161172084, "epoch": 0.6667342182591954, "step": 1645}, {"loss": 0.3476806640625, "grad_norm": 0.4700915515422821, "learning_rate": 2.6164659096812285e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.88284849865816, "epoch": 0.6687607660350593, "step": 1650}, {"loss": 0.323876953125, "grad_norm": 0.3228314518928528, "learning_rate": 2.5876660693710994e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8916923522350968, "epoch": 0.670787313810923, "step": 1655}, {"loss": 0.3321533203125, "grad_norm": 0.4141518771648407, "learning_rate": 2.5589701708584526e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8889392826343863, "epoch": 0.6728138615867869, "step": 1660}, {"loss": 0.3514404296875, "grad_norm": 0.5032157897949219, "learning_rate": 2.5303794505820863e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8822658696270993, "epoch": 0.6748404093626508, "step": 1665}, {"loss": 0.3413818359375, "grad_norm": 0.5460382699966431, "learning_rate": 2.501895140448919e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8864838250847952, "epoch": 0.6768669571385145, "step": 1670}, {"loss": 0.323388671875, "grad_norm": 0.4051891565322876, "learning_rate": 2.4735184677809054e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8904025200370893, "epoch": 0.6788935049143784, "step": 1675}, {"loss": 0.3535400390625, "grad_norm": 0.4227135479450226, "learning_rate": 2.4452506552621575e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8815619943211309, "epoch": 0.6809200526902421, "step": 1680}, {"loss": 0.34765625, "grad_norm": 0.4865545630455017, "learning_rate": 2.4170929208862525e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8828536786093468, "epoch": 0.682946600466106, "step": 1685}, {"loss": 0.3500244140625, "grad_norm": 0.4662053883075714, "learning_rate": 2.3890464779037658e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8823926432126887, "epoch": 0.6849731482419698, "step": 1690}, {"loss": 0.339892578125, "grad_norm": 0.5017711520195007, "learning_rate": 2.361112534769986e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8856021181411485, "epoch": 0.6869996960178336, "step": 1695}, {"loss": 0.354833984375, "grad_norm": 0.42811140418052673, "learning_rate": 2.3332922950928495e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8817200279273888, "epoch": 0.6890262437936975, "step": 1700}, {"loss": 0.3400634765625, "grad_norm": 0.45597004890441895, "learning_rate": 2.305586957581073e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8865903430650272, "epoch": 0.6910527915695612, "step": 1705}, {"loss": 0.32783203125, "grad_norm": 0.4592193365097046, "learning_rate": 2.2779977159925137e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8893759222289732, "epoch": 0.6930793393454251, "step": 1710}, {"loss": 0.3681396484375, "grad_norm": 0.5005931258201599, "learning_rate": 2.250525759082728e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8778964980174249, "epoch": 0.6951058871212888, "step": 1715}, {"loss": 0.292578125, "grad_norm": 0.2610883414745331, "learning_rate": 2.2231722705537513e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.9055491418456092, "epoch": 0.6971324348971527, "step": 1720}, {"loss": 0.35205078125, "grad_norm": 0.4703691601753235, "learning_rate": 2.1959384290030912e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8827783713387997, "epoch": 0.6991589826730166, "step": 1725}, {"loss": 0.32626953125, "grad_norm": 0.3500620126724243, "learning_rate": 2.1688254078729524e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8898699068503675, "epoch": 0.7011855304488803, "step": 1730}, {"loss": 0.358837890625, "grad_norm": 0.3912479877471924, "learning_rate": 2.141834375399671e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8790342914775593, "epoch": 0.7032120782247442, "step": 1735}, {"loss": 0.3556884765625, "grad_norm": 0.4508066475391388, "learning_rate": 2.1149664945633797e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.880684688811495, "epoch": 0.7052386260006079, "step": 1740}, {"loss": 0.3451904296875, "grad_norm": 0.3865109384059906, "learning_rate": 2.0882229230378916e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8843628924774509, "epoch": 0.7072651737764718, "step": 1745}, {"loss": 0.302392578125, "grad_norm": 0.46912604570388794, "learning_rate": 2.0616048131408297e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.900628045174191, "epoch": 0.7092917215523356, "step": 1750}, {"loss": 0.34765625, "grad_norm": 0.4662412703037262, "learning_rate": 2.035113311783967e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.882717010509926, "epoch": 0.7113182693281994, "step": 1755}, {"loss": 0.3470947265625, "grad_norm": 0.4677916467189789, "learning_rate": 2.0087495604238165e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8837272721507348, "epoch": 0.7133448171040633, "step": 1760}, {"loss": 0.3470703125, "grad_norm": 0.5265308618545532, "learning_rate": 1.9825146950124367e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.882448854961832, "epoch": 0.715371364879927, "step": 1765}, {"loss": 0.3476318359375, "grad_norm": 0.5809183716773987, "learning_rate": 1.9564098459485003e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8854824391402744, "epoch": 0.7173979126557909, "step": 1770}, {"loss": 0.341748046875, "grad_norm": 0.44178304076194763, "learning_rate": 1.930436138028578e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8857181884786847, "epoch": 0.7194244604316546, "step": 1775}, {"loss": 0.3384765625, "grad_norm": 0.3675595819950104, "learning_rate": 1.9045946903986801e-06, "tIoU": 0.9989814758300781, "tRecall_30": 0.9996322631835938, "tRecall_50": 0.9996322631835938, "tRecall_70": 0.9996322631835938, "token_acc": 0.8861089376698108, "epoch": 0.7214510082075185, "step": 1780}, {"loss": 0.3569091796875, "grad_norm": 0.47204509377479553, "learning_rate": 1.878886616506026e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8793076633963633, "epoch": 0.7234775559833823, "step": 1785}, {"loss": 0.3317626953125, "grad_norm": 0.3641361892223358, "learning_rate": 1.85331302405108e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8881008958559753, "epoch": 0.7255041037592461, "step": 1790}, {"loss": 0.3068359375, "grad_norm": 0.4817829728126526, "learning_rate": 1.827875014939815e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8976465497883155, "epoch": 0.72753065153511, "step": 1795}, {"loss": 0.3442626953125, "grad_norm": 0.4408557116985321, "learning_rate": 1.802573685236239e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8833064053786418, "epoch": 0.7295571993109737, "step": 1800}, {"loss": 0.311279296875, "grad_norm": 0.4439926743507385, "learning_rate": 1.77741012511516e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8974859360066844, "epoch": 0.7315837470868376, "step": 1805}, {"loss": 0.3204345703125, "grad_norm": 0.4358379542827606, "learning_rate": 1.7523854188152235e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8913284952606635, "epoch": 0.7336102948627014, "step": 1810}, {"loss": 0.3365966796875, "grad_norm": 0.36242666840553284, "learning_rate": 1.7275006445921889e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8881554550174293, "epoch": 0.7356368426385652, "step": 1815}, {"loss": 0.3471435546875, "grad_norm": 0.4304690659046173, "learning_rate": 1.702756874672473e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8829572785903128, "epoch": 0.737663390414429, "step": 1820}, {"loss": 0.3568359375, "grad_norm": 0.40860116481781006, "learning_rate": 1.6781551752069441e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8813770378742888, "epoch": 0.7396899381902928, "step": 1825}, {"loss": 0.324951171875, "grad_norm": 0.572250485420227, "learning_rate": 1.6536966062249932e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.889588998430743, "epoch": 0.7417164859661567, "step": 1830}, {"loss": 0.3470703125, "grad_norm": 0.42056840658187866, "learning_rate": 1.6293822215888533e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8832219765723289, "epoch": 0.7437430337420204, "step": 1835}, {"loss": 0.365869140625, "grad_norm": 0.45901891589164734, "learning_rate": 1.6052130689481948e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8776322582398174, "epoch": 0.7457695815178843, "step": 1840}, {"loss": 0.3380126953125, "grad_norm": 0.4727613627910614, "learning_rate": 1.5811901896949788e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8879822377210093, "epoch": 0.7477961292937481, "step": 1845}, {"loss": 0.3182373046875, "grad_norm": 0.24706824123859406, "learning_rate": 1.5573146189185945e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8955709839597797, "epoch": 0.7498226770696119, "step": 1850}, {"loss": 0.3082763671875, "grad_norm": 0.41805166006088257, "learning_rate": 1.533587385361252e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8968251179894365, "epoch": 0.7518492248454758, "step": 1855}, {"loss": 0.324169921875, "grad_norm": 0.38584384322166443, "learning_rate": 1.5100095113736635e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8908545262919716, "epoch": 0.7538757726213395, "step": 1860}, {"loss": 0.34052734375, "grad_norm": 0.4495343863964081, "learning_rate": 1.4865820128709812e-06, "tIoU": 0.99375, "tRecall_30": 0.99375, "tRecall_50": 0.99375, "tRecall_70": 0.99375, "token_acc": 0.8863164086726897, "epoch": 0.7559023203972034, "step": 1865}, {"loss": 0.3251708984375, "grad_norm": 0.40899696946144104, "learning_rate": 1.4633058992890387e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8894297164829809, "epoch": 0.7579288681730671, "step": 1870}, {"loss": 0.34287109375, "grad_norm": 0.30098530650138855, "learning_rate": 1.4401821735408456e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8844233449875459, "epoch": 0.759955415948931, "step": 1875}, {"loss": 0.297802734375, "grad_norm": 0.392616868019104, "learning_rate": 1.4172118319733796e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.9007301045196695, "epoch": 0.7619819637247948, "step": 1880}, {"loss": 0.341845703125, "grad_norm": 0.46458375453948975, "learning_rate": 1.3943958643246514e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8854414186853464, "epoch": 0.7640085115006586, "step": 1885}, {"loss": 0.3319580078125, "grad_norm": 0.4317547082901001, "learning_rate": 1.3717352536810679e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8870785499809603, "epoch": 0.7660350592765225, "step": 1890}, {"loss": 0.337744140625, "grad_norm": 0.48436179757118225, "learning_rate": 1.3492309764350597e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8858699511332313, "epoch": 0.7680616070523862, "step": 1895}, {"loss": 0.3273193359375, "grad_norm": 0.41826820373535156, "learning_rate": 1.3268840022430306e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.890079424999245, "epoch": 0.7700881548282501, "step": 1900}, {"loss": 0.350390625, "grad_norm": 0.5227497220039368, "learning_rate": 1.3046952939835545e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8822442209117562, "epoch": 0.7721147026041139, "step": 1905}, {"loss": 0.328564453125, "grad_norm": 0.412296861410141, "learning_rate": 1.2826658077159054e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.888492897475502, "epoch": 0.7741412503799777, "step": 1910}, {"loss": 0.32509765625, "grad_norm": 0.47801196575164795, "learning_rate": 1.2607964926388494e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8916572717023675, "epoch": 0.7761677981558415, "step": 1915}, {"loss": 0.325634765625, "grad_norm": 0.41708609461784363, "learning_rate": 1.2390882910497627e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8909982773219656, "epoch": 0.7781943459317053, "step": 1920}, {"loss": 0.3413818359375, "grad_norm": 0.39932745695114136, "learning_rate": 1.2175421383040088e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8852147296169186, "epoch": 0.7802208937075692, "step": 1925}, {"loss": 0.3327392578125, "grad_norm": 0.3740224838256836, "learning_rate": 1.1961589627746566e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.888258064516129, "epoch": 0.782247441483433, "step": 1930}, {"loss": 0.3427490234375, "grad_norm": 0.4308731257915497, "learning_rate": 1.1749396858124611e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8836409967702746, "epoch": 0.7842739892592968, "step": 1935}, {"loss": 0.34677734375, "grad_norm": 0.43539366126060486, "learning_rate": 1.1538852217061853e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8837174094771694, "epoch": 0.7863005370351606, "step": 1940}, {"loss": 0.312109375, "grad_norm": 0.40453237295150757, "learning_rate": 1.1329964776431812e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8960424425305664, "epoch": 0.7883270848110244, "step": 1945}, {"loss": 0.3400146484375, "grad_norm": 0.46798184514045715, "learning_rate": 1.112274353670323e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8842769195905881, "epoch": 0.7903536325868883, "step": 1950}, {"loss": 0.33603515625, "grad_norm": 0.5182797908782959, "learning_rate": 1.0917197426552089e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8882436625418512, "epoch": 0.792380180362752, "step": 1955}, {"loss": 0.3489990234375, "grad_norm": 0.40419793128967285, "learning_rate": 1.0713335302477073e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.882445557326049, "epoch": 0.7944067281386159, "step": 1960}, {"loss": 0.351708984375, "grad_norm": 0.47690391540527344, "learning_rate": 1.0511165948417774e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8830510993235429, "epoch": 0.7964332759144797, "step": 1965}, {"loss": 0.30078125, "grad_norm": 0.35004881024360657, "learning_rate": 1.0310698075376368e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.9015132509236875, "epoch": 0.7984598236903435, "step": 1970}, {"loss": 0.3324951171875, "grad_norm": 0.43832194805145264, "learning_rate": 1.0111940321042146e-06, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8875856153256306, "epoch": 0.8004863714662073, "step": 1975}, {"loss": 0.3274658203125, "grad_norm": 0.33005771040916443, "learning_rate": 9.914901249419512e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8900861560250206, "epoch": 0.8025129192420711, "step": 1980}, {"loss": 0.3199951171875, "grad_norm": 0.2617853581905365, "learning_rate": 9.719589350458753e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8958106699301968, "epoch": 0.804539467017935, "step": 1985}, {"loss": 0.3506591796875, "grad_norm": 0.5502761602401733, "learning_rate": 9.526013039690429e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.880904348734421, "epoch": 0.8065660147937987, "step": 1990}, {"loss": 0.3437744140625, "grad_norm": 0.4234298765659332, "learning_rate": 9.334180657862629e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8829914034772302, "epoch": 0.8085925625696626, "step": 1995}, {"loss": 0.3411865234375, "grad_norm": 0.4022880494594574, "learning_rate": 9.144100470581718e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.884309301945398, "epoch": 0.8106191103455264, "step": 2000}, {"loss": 0.338232421875, "grad_norm": 0.41832342743873596, "learning_rate": 8.955780667956032e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8855404089581305, "epoch": 0.8126456581213902, "step": 2005}, {"loss": 0.3219970703125, "grad_norm": 0.4092077612876892, "learning_rate": 8.769229364243137e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.891176808231676, "epoch": 0.814672205897254, "step": 2010}, {"loss": 0.3410888671875, "grad_norm": 0.4354773759841919, "learning_rate": 8.584454597500092e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8857620386816657, "epoch": 0.8166987536731178, "step": 2015}, {"loss": 0.3350830078125, "grad_norm": 0.49091804027557373, "learning_rate": 8.401464329237185e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8864969363347422, "epoch": 0.8187253014489817, "step": 2020}, {"loss": 0.3209716796875, "grad_norm": 0.4098585546016693, "learning_rate": 8.220266444074853e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8926854374810778, "epoch": 0.8207518492248455, "step": 2025}, {"loss": 0.314892578125, "grad_norm": 0.3522760272026062, "learning_rate": 8.040868749403968e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8934764299338601, "epoch": 0.8227783970007093, "step": 2030}, {"loss": 0.3139892578125, "grad_norm": 0.40749940276145935, "learning_rate": 7.863278975049388e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8930227973747072, "epoch": 0.8248049447765731, "step": 2035}, {"loss": 0.3261474609375, "grad_norm": 0.4043844938278198, "learning_rate": 7.687504772936983e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8905281649048885, "epoch": 0.8268314925524369, "step": 2040}, {"loss": 0.3332763671875, "grad_norm": 0.4920687973499298, "learning_rate": 7.513553716763849e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8874587937196271, "epoch": 0.8288580403283008, "step": 2045}, {"loss": 0.2773193359375, "grad_norm": 0.32292449474334717, "learning_rate": 7.341433301672047e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.9084233121903194, "epoch": 0.8308845881041645, "step": 2050}, {"loss": 0.3315673828125, "grad_norm": 0.3950773775577545, "learning_rate": 7.171150943925565e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8875533463255791, "epoch": 0.8329111358800284, "step": 2055}, {"loss": 0.340283203125, "grad_norm": 0.43536999821662903, "learning_rate": 7.002713980590864e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8850941570345794, "epoch": 0.8349376836558922, "step": 2060}, {"loss": 0.3458251953125, "grad_norm": 0.47620904445648193, "learning_rate": 6.836129669220687e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8832007379222876, "epoch": 0.836964231431756, "step": 2065}, {"loss": 0.32021484375, "grad_norm": 0.32710176706314087, "learning_rate": 6.671405187541369e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8921389175833802, "epoch": 0.8389907792076198, "step": 2070}, {"loss": 0.319482421875, "grad_norm": 0.3626502752304077, "learning_rate": 6.508547633143519e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8915977638033111, "epoch": 0.8410173269834836, "step": 2075}, {"loss": 0.3444580078125, "grad_norm": 0.4492169916629791, "learning_rate": 6.347564023176283e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8843687264694935, "epoch": 0.8430438747593475, "step": 2080}, {"loss": 0.3422119140625, "grad_norm": 0.4278513193130493, "learning_rate": 6.188461294044923e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8840360895664138, "epoch": 0.8450704225352113, "step": 2085}, {"loss": 0.347705078125, "grad_norm": 0.4415869414806366, "learning_rate": 6.031246301111987e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8835707624529697, "epoch": 0.8470969703110751, "step": 2090}, {"loss": 0.3265625, "grad_norm": 0.3770047724246979, "learning_rate": 5.875925818401873e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8900356756694248, "epoch": 0.8491235180869389, "step": 2095}, {"loss": 0.3359619140625, "grad_norm": 0.37200838327407837, "learning_rate": 5.722506538309025e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8867193743928979, "epoch": 0.8511500658628027, "step": 2100}, {"loss": 0.346875, "grad_norm": 0.4725235402584076, "learning_rate": 5.570995071309515e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8840582633957392, "epoch": 0.8531766136386665, "step": 2105}, {"loss": 0.32822265625, "grad_norm": 0.3805801272392273, "learning_rate": 5.421397945676255e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8886891705894553, "epoch": 0.8552031614145303, "step": 2110}, {"loss": 0.3368408203125, "grad_norm": 0.44028645753860474, "learning_rate": 5.273721607197651e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8858809924306139, "epoch": 0.8572297091903942, "step": 2115}, {"loss": 0.3289794921875, "grad_norm": 0.3674177825450897, "learning_rate": 5.127972418899941e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8882873734649193, "epoch": 0.859256256966258, "step": 2120}, {"loss": 0.3440185546875, "grad_norm": 0.4421343505382538, "learning_rate": 4.984156660772976e-07, "tIoU": 0.9989355087280274, "tRecall_30": 0.9997768402099609, "tRecall_50": 0.9997768402099609, "tRecall_70": 0.9997768402099609, "token_acc": 0.883937518420277, "epoch": 0.8612828047421218, "step": 2125}, {"loss": 0.351904296875, "grad_norm": 0.43405815958976746, "learning_rate": 4.842280529499671e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8818160076530612, "epoch": 0.8633093525179856, "step": 2130}, {"loss": 0.30791015625, "grad_norm": 0.37876492738723755, "learning_rate": 4.7023501381889326e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8961273254511662, "epoch": 0.8653359002938494, "step": 2135}, {"loss": 0.3364501953125, "grad_norm": 0.425828218460083, "learning_rate": 4.5643715161123424e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8863891502546964, "epoch": 0.8673624480697133, "step": 2140}, {"loss": 0.34580078125, "grad_norm": 0.4508116841316223, "learning_rate": 4.428350608444337e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8834513383656141, "epoch": 0.869388995845577, "step": 2145}, {"loss": 0.3091552734375, "grad_norm": 0.3458070158958435, "learning_rate": 4.2942932760060395e-07, "tIoU": 0.9999964714050293, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8950527958034684, "epoch": 0.8714155436214409, "step": 2150}, {"loss": 0.35849609375, "grad_norm": 0.5336869359016418, "learning_rate": 4.1622052950126946e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8784318039617439, "epoch": 0.8734420913973047, "step": 2155}, {"loss": 0.3381591796875, "grad_norm": 0.35457611083984375, "learning_rate": 4.0320923568248796e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8862457701347886, "epoch": 0.8754686391731685, "step": 2160}, {"loss": 0.3113525390625, "grad_norm": 0.3743188679218292, "learning_rate": 3.903960067703177e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8935743921322947, "epoch": 0.8774951869490323, "step": 2165}, {"loss": 0.353515625, "grad_norm": 0.4452930688858032, "learning_rate": 3.7778139485667055e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8812360435248409, "epoch": 0.8795217347248961, "step": 2170}, {"loss": 0.3264892578125, "grad_norm": 0.4194066524505615, "learning_rate": 3.653659434755125e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8891549098383631, "epoch": 0.88154828250076, "step": 2175}, {"loss": 0.3505615234375, "grad_norm": 0.4035830497741699, "learning_rate": 3.531501875794552e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8810427993168125, "epoch": 0.8835748302766238, "step": 2180}, {"loss": 0.345849609375, "grad_norm": 0.37540069222450256, "learning_rate": 3.411346535166998e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8822157717672492, "epoch": 0.8856013780524876, "step": 2185}, {"loss": 0.33671875, "grad_norm": 0.4143151044845581, "learning_rate": 3.293198590083613e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8863002774610664, "epoch": 0.8876279258283514, "step": 2190}, {"loss": 0.3336669921875, "grad_norm": 0.44223761558532715, "learning_rate": 3.1770631312615483e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8871216076018793, "epoch": 0.8896544736042152, "step": 2195}, {"loss": 0.312890625, "grad_norm": 0.3474159836769104, "learning_rate": 3.0629451627047e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8945662581101734, "epoch": 0.891681021380079, "step": 2200}, {"loss": 0.328857421875, "grad_norm": 0.3162406086921692, "learning_rate": 2.9508496014880205e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8887455197132617, "epoch": 0.8937075691559428, "step": 2205}, {"loss": 0.3455322265625, "grad_norm": 0.4361012279987335, "learning_rate": 2.840781277545712e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.885535628989535, "epoch": 0.8957341169318067, "step": 2210}, {"loss": 0.3302978515625, "grad_norm": 0.41882529854774475, "learning_rate": 2.732744933463055e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8884911839624202, "epoch": 0.8977606647076705, "step": 2215}, {"loss": 0.3207763671875, "grad_norm": 0.48173168301582336, "learning_rate": 2.626745224272126e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8912699949407569, "epoch": 0.8997872124835343, "step": 2220}, {"loss": 0.33603515625, "grad_norm": 0.3567911982536316, "learning_rate": 2.5227867172511746e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8863890027430085, "epoch": 0.9018137602593981, "step": 2225}, {"loss": 0.303759765625, "grad_norm": 0.4256726801395416, "learning_rate": 2.4208738917278654e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8972713804085347, "epoch": 0.9038403080352619, "step": 2230}, {"loss": 0.34375, "grad_norm": 0.49550139904022217, "learning_rate": 2.3210111388862312e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8858439106262457, "epoch": 0.9058668558111257, "step": 2235}, {"loss": 0.3441650390625, "grad_norm": 0.38924527168273926, "learning_rate": 2.2232027615775088e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.884348966874385, "epoch": 0.9078934035869896, "step": 2240}, {"loss": 0.32607421875, "grad_norm": 0.38762596249580383, "learning_rate": 2.127452974134725e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8882261345656378, "epoch": 0.9099199513628534, "step": 2245}, {"loss": 0.3317138671875, "grad_norm": 0.520622193813324, "learning_rate": 2.0337659021911094e-07, "tIoU": 0.9982122421264649, "tRecall_30": 0.999609375, "tRecall_50": 0.999609375, "tRecall_70": 0.999609375, "token_acc": 0.8874109294771036, "epoch": 0.9119464991387172, "step": 2250}, {"loss": 0.3436767578125, "grad_norm": 0.5936887264251709, "learning_rate": 1.9421455825023194e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8838248321372174, "epoch": 0.913973046914581, "step": 2255}, {"loss": 0.3408447265625, "grad_norm": 0.526875376701355, "learning_rate": 1.8525959627725188e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8860013722851187, "epoch": 0.9159995946904448, "step": 2260}, {"loss": 0.3270751953125, "grad_norm": 0.444475382566452, "learning_rate": 1.7651209014842963e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8931827257687261, "epoch": 0.9180261424663086, "step": 2265}, {"loss": 0.3276611328125, "grad_norm": 0.34051868319511414, "learning_rate": 1.6797241677323773e-07, "tIoU": 0.9999456405639648, "tRecall_30": 0.9999431610107422, "tRecall_50": 0.9999431610107422, "tRecall_70": 0.9999431610107422, "token_acc": 0.8906207714215756, "epoch": 0.9200526902421725, "step": 2270}, {"loss": 0.3405029296875, "grad_norm": 0.40930411219596863, "learning_rate": 1.5964094410612418e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8846896342196707, "epoch": 0.9220792380180363, "step": 2275}, {"loss": 0.3163818359375, "grad_norm": 0.35660627484321594, "learning_rate": 1.5151803113065854e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.893365429779239, "epoch": 0.9241057857939001, "step": 2280}, {"loss": 0.3478759765625, "grad_norm": 0.48520004749298096, "learning_rate": 1.4360402784406202e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8817491056381752, "epoch": 0.9261323335697639, "step": 2285}, {"loss": 0.3416748046875, "grad_norm": 0.4974427819252014, "learning_rate": 1.3589927524213232e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8828711924600464, "epoch": 0.9281588813456277, "step": 2290}, {"loss": 0.3143310546875, "grad_norm": 0.4171598553657532, "learning_rate": 1.2840410530454262e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8934846058623783, "epoch": 0.9301854291214915, "step": 2295}, {"loss": 0.3171875, "grad_norm": 0.3226058781147003, "learning_rate": 1.211188409805447e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8943094746978785, "epoch": 0.9322119768973554, "step": 2300}, {"loss": 0.318310546875, "grad_norm": 0.443516343832016, "learning_rate": 1.1404379617504846e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.892321063327866, "epoch": 0.9342385246732192, "step": 2305}, {"loss": 0.331787109375, "grad_norm": 0.45417943596839905, "learning_rate": 1.0717927573510312e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8895140322354761, "epoch": 0.936265072449083, "step": 2310}, {"loss": 0.33115234375, "grad_norm": 0.4532037675380707, "learning_rate": 1.0052557543675346e-07, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8877487976546694, "epoch": 0.9382916202249468, "step": 2315}, {"loss": 0.318115234375, "grad_norm": 0.43640556931495667, "learning_rate": 9.40829819723027e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.894122094364284, "epoch": 0.9403181680008106, "step": 2320}, {"loss": 0.3271728515625, "grad_norm": 0.39802286028862, "learning_rate": 8.785177293795577e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8903346784666747, "epoch": 0.9423447157766744, "step": 2325}, {"loss": 0.31474609375, "grad_norm": 0.35078153014183044, "learning_rate": 8.183221682186048e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8940282315868322, "epoch": 0.9443712635525382, "step": 2330}, {"loss": 0.33984375, "grad_norm": 0.4221342206001282, "learning_rate": 7.602457299253685e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8851874961003307, "epoch": 0.9463978113284021, "step": 2335}, {"loss": 0.3451416015625, "grad_norm": 0.46918657422065735, "learning_rate": 7.042909168770373e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.884408887239176, "epoch": 0.9484243591042659, "step": 2340}, {"loss": 0.3374267578125, "grad_norm": 0.2663903534412384, "learning_rate": 6.504601400349419e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8878492824179726, "epoch": 0.9504509068801297, "step": 2345}, {"loss": 0.319580078125, "grad_norm": 0.43997228145599365, "learning_rate": 5.987557188407156e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8926250406178918, "epoch": 0.9524774546559935, "step": 2350}, {"loss": 0.33017578125, "grad_norm": 0.37927335500717163, "learning_rate": 5.491798811162852e-08, "tIoU": 0.9998649597167969, "tRecall_30": 0.9998437881469726, "tRecall_50": 0.9998437881469726, "tRecall_70": 0.9998437881469726, "token_acc": 0.8897617369237276, "epoch": 0.9545040024318573, "step": 2355}, {"loss": 0.3291748046875, "grad_norm": 0.4116072654724121, "learning_rate": 5.0173476296795363e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8886311682565646, "epoch": 0.9565305502077212, "step": 2360}, {"loss": 0.343798828125, "grad_norm": 0.4057512581348419, "learning_rate": 4.564224086943014e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8851102244892806, "epoch": 0.958557097983585, "step": 2365}, {"loss": 0.3299072265625, "grad_norm": 0.4532250463962555, "learning_rate": 4.132447706981624e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.889131115705068, "epoch": 0.9605836457594488, "step": 2370}, {"loss": 0.315185546875, "grad_norm": 0.3738594651222229, "learning_rate": 3.722037094024244e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8929597656371739, "epoch": 0.9626101935353126, "step": 2375}, {"loss": 0.330908203125, "grad_norm": 0.3902566432952881, "learning_rate": 3.333009931699438e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8876850592640813, "epoch": 0.9646367413111764, "step": 2380}, {"loss": 0.3281982421875, "grad_norm": 0.4984258711338043, "learning_rate": 2.9653829822728353e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8927605491760068, "epoch": 0.9666632890870402, "step": 2385}, {"loss": 0.317041015625, "grad_norm": 0.3534846305847168, "learning_rate": 2.6191720859254922e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8924771605711797, "epoch": 0.968689836862904, "step": 2390}, {"loss": 0.3455322265625, "grad_norm": 0.4049256145954132, "learning_rate": 2.2943921600708796e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8843115851136679, "epoch": 0.9707163846387679, "step": 2395}, {"loss": 0.3220458984375, "grad_norm": 0.3752588927745819, "learning_rate": 1.9910571987125094e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8901173485569299, "epoch": 0.9727429324146317, "step": 2400}, {"loss": 0.329150390625, "grad_norm": 0.4580046534538269, "learning_rate": 1.709180271840638e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8885664172418579, "epoch": 0.9747694801904955, "step": 2405}, {"loss": 0.3216552734375, "grad_norm": 0.35217657685279846, "learning_rate": 1.4487735248692737e-08, "tIoU": 0.9999425888061524, "tRecall_30": 0.9999431610107422, "tRecall_50": 0.9999431610107422, "tRecall_70": 0.9999431610107422, "token_acc": 0.8933333333333333, "epoch": 0.9767960279663593, "step": 2410}, {"loss": 0.3031494140625, "grad_norm": 0.41408440470695496, "learning_rate": 1.2098481781130377e-08, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8987533738838125, "epoch": 0.9788225757422231, "step": 2415}, {"loss": 0.3330322265625, "grad_norm": 0.34325486421585083, "learning_rate": 9.924145263033313e-09, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.887853635954646, "epoch": 0.980849123518087, "step": 2420}, {"loss": 0.310400390625, "grad_norm": 0.379108190536499, "learning_rate": 7.964819381449662e-09, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8947683743455007, "epoch": 0.9828756712939507, "step": 2425}, {"loss": 0.3291748046875, "grad_norm": 0.46881887316703796, "learning_rate": 6.22058855912544e-09, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.889718241829013, "epoch": 0.9849022190698146, "step": 2430}, {"loss": 0.3313232421875, "grad_norm": 0.38360169529914856, "learning_rate": 4.691527950865804e-09, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8881468006851406, "epoch": 0.9869287668456784, "step": 2435}, {"loss": 0.3405029296875, "grad_norm": 0.5358332395553589, "learning_rate": 3.3777034402981966e-09, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8851056035571725, "epoch": 0.9889553146215422, "step": 2440}, {"loss": 0.3314453125, "grad_norm": 0.3311910033226013, "learning_rate": 2.2791716370307305e-09, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8888736476734616, "epoch": 0.990981862397406, "step": 2445}, {"loss": 0.3162109375, "grad_norm": 0.4242856800556183, "learning_rate": 1.3959798742169128e-09, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8925040315518001, "epoch": 0.9930084101732698, "step": 2450}, {"loss": 0.334423828125, "grad_norm": 0.49177971482276917, "learning_rate": 7.28166206512837e-10, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8867867609032926, "epoch": 0.9950349579491337, "step": 2455}, {"loss": 0.3192626953125, "grad_norm": 0.5094365477561951, "learning_rate": 2.7575940843960204e-10, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8922791595741978, "epoch": 0.9970615057249975, "step": 2460}, {"loss": 0.3603271484375, "grad_norm": 0.3633192181587219, "learning_rate": 3.877897314263912e-11, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8784863668900892, "epoch": 0.9990880535008613, "step": 2465}, {"train_runtime": 35960.2414, "train_samples_per_second": 2.196, "train_steps_per_second": 0.069, "total_flos": 8.412078391672858e+19, "train_loss": 0.36147492592686387, "tIoU": 1.0, "tRecall_30": 1.0, "tRecall_50": 1.0, "tRecall_70": 1.0, "token_acc": 0.8868830792044099, "epoch": 1.0, "step": 2468}], "memory": 55.82421875} diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/args.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/args.json new file mode 100644 index 0000000000000000000000000000000000000000..a01674fc2ac677e1f85b57fb7da2e12d76c3b2f0 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/args.json @@ -0,0 +1,706 @@ +{ + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "checkpoints/Qwen3-VL-30B-A3B-Instruct", + "model_type": "qwen3_moe_vl", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": null, + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_type": "default" + }, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_vl", + "system": null, + "max_length": 16384, + "truncation_strategy": "delete", + "max_pixels": null, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_side": "right", + "padding_free": true, + "loss_scale": "default", + "sequence_parallel_size": 1, + "template_backend": "swift", + "response_prefix": null, + "enable_thinking": null, + "add_non_thinking_prefix": true, + "dataset": [ + "../data/cosmo-cap/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/moment10m/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/vtimellm/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/timelens-100k/train/train_w_cot_reasoning_32k.jsonl" + ], + "val_dataset": [], + "cached_dataset": [], + "cached_val_dataset": [], + "split_dataset_ratio": 0.0, + "data_seed": 42, + "dataset_num_proc": 8, + "load_from_cache_file": false, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "remove_unused_columns": true, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": null, + "temperature": 0.9, + "top_k": 50, + "top_p": 0.9, + "repetition_penalty": 1.0, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "structured_outputs_regex": null, + "ckpt_dir": null, + "lora_modules": [], + "tuner_backend": "peft", + "train_type": "lora", + "adapters": [], + "external_plugins": [], + "seed": 42, + "model_kwargs": {}, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "custom_register_path": [], + "use_hf": false, + "hub_token": null, + "ddp_timeout": 18000000, + "ddp_backend": null, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": true, + "freeze_parameters": [], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [], + "trainable_parameters_regex": null, + "adapter_load": null, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "use_rslora": false, + "rlhf_type": null, + "ref_load": null, + "ref_adapter_load": null, + "beta": null, + "rpo_alpha": null, + "reference_free": false, + "label_smoothing": 0.0, + "f_divergence_type": "reverse_kl", + "loss_type": null, + "desirable_weight": 1.0, + "undesirable_weight": 1.0, + "calculate_KL": null, + "center_rewards_coefficient": null, + "teacher_model": null, + "teacher_model_type": null, + "teacher_model_revision": null, + "lmbda": 0.5, + "seq_kd": false, + "offload_teacher_model": false, + "sft_alpha": 0.0, + "generation_batch_size": null, + "steps_per_generation": null, + "num_generations": 8, + "num_generations_eval": null, + "max_completion_length": 512, + "importance_sampling_level": "token", + "tau_pos": 1.0, + "tau_neg": 1.05, + "epsilon": 0.2, + "epsilon_high": null, + "delta": null, + "use_vllm": true, + "vllm_mode": null, + "vllm_enable_prefix_caching": true, + "vllm_gpu_memory_utilization": 0.9, + "vllm_tensor_parallel_size": 1, + "vllm_max_model_len": null, + "vllm_enforce_eager": false, + "vllm_limit_mm_per_prompt": null, + "vllm_disable_cascade_attn": false, + "vllm_max_num_seqs": null, + "vllm_mm_processor_cache_gb": null, + "vllm_engine_kwargs": null, + "sleep_level": 0, + "offload_optimizer": false, + "offload_model": false, + "offload_bridge": false, + "vllm_server_base_url": null, + "vllm_server_host": null, + "vllm_server_port": [ + 8000 + ], + "vllm_server_timeout": 240.0, + "vllm_server_group_port": null, + "reward_funcs": [], + "reward_weights": null, + "cosine_min_len_value_wrong": -0.5, + "cosine_max_len_value_wrong": 0.0, + "cosine_min_len_value_correct": 1.0, + "cosine_max_len_value_correct": 0.5, + "cosine_max_len": null, + "repetition_n_grams": 3, + "repetition_max_penalty": -1.0, + "soft_max_length": null, + "soft_cache_length": null, + "dynamic_sample": false, + "max_resample_times": 3, + "overlong_filter": false, + "scale_rewards": "group", + "advantage_estimator": "grpo", + "kl_in_reward": false, + "wandb_log_unique_prompts": null, + "log_completions": false, + "rollout_importance_sampling_mode": null, + "rollout_importance_sampling_threshold": 2.0, + "log_rollout_offpolicy_metrics": false, + "off_policy_sequence_mask_delta": null, + "log_entropy": false, + "top_entropy_quantile": 1.0, + "reward_model": null, + "reward_model_plugin": null, + "sync_ref_model": false, + "ref_model_sync_steps": 512, + "ref_model_mixup_alpha": 0.6, + "async_generate": false, + "move_model_batches": null, + "multi_turn_scheduler": null, + "max_turns": null, + "completion_length_limit_scope": "per_round", + "vllm_server_pass_dataset": false, + "num_iterations": 1, + "check_model": true, + "padded_vocab_size": 151936, + "initialize_embedding": false, + "mlp_padding_free": false, + "load_safetensors": true, + "save_safetensors": true, + "ref_model": null, + "ref_adapters": [], + "merge_lora": true, + "max_shard_size": "5GB", + "train_dataloader_shuffle": true, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": true, + "dataloader_prefetch_factor": 2, + "group_by_length": false, + "architectures": "Qwen3VLMoeForConditionalGeneration", + "llm_architectures": "Qwen3VLMoeForConditionalGeneration", + "max_epochs": 1, + "enable_dft_loss": false, + "enable_channel_loss": false, + "save_strategy": "steps", + "original_max_position_embeddings": null, + "partial_rotary_factor": null, + "use_shared_expert_gate": false, + "report_to": "wandb", + "vit_gradient_checkpointing": true, + "vit_lr": null, + "aligner_lr": null, + "gradient_checkpointing_kwargs": null, + "linear_num_value_heads": null, + "linear_num_key_heads": null, + "linear_key_head_dim": null, + "linear_value_head_dim": null, + "linear_conv_kernel_dim": null, + "layer_types": null, + "apply_wd_to_qk_layernorm": false, + "mrope_interleaved": true, + "micro_batch_size": 12, + "global_batch_size": 96, + "recompute_granularity": "full", + "recompute_method": "uniform", + "recompute_num_layers": 1, + "recompute_modules": [ + "core_attn" + ], + "use_cpu_initialization": false, + "deterministic_mode": false, + "train_iters": null, + "log_interval": 5, + "tensorboard_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs", + "no_masked_softmax_fusion": false, + "no_bias_dropout_fusion": false, + "no_bias_swiglu_fusion": false, + "no_rope_fusion": false, + "no_gradient_accumulation_fusion": true, + "cross_entropy_loss_fusion": true, + "cross_entropy_fusion_impl": "native", + "calculate_per_token_loss": true, + "use_flash_attn": false, + "attention_backend": "flash", + "optimizer": "adam", + "optimizer_cpu_offload": false, + "optimizer_offload_fraction": 1.0, + "use_precision_aware_optimizer": false, + "main_grads_dtype": "fp32", + "main_params_dtype": "fp32", + "exp_avg_dtype": "fp32", + "exp_avg_sq_dtype": "fp32", + "dataloader_type": "cyclic", + "manual_gc": false, + "manual_gc_interval": 0, + "lr": 1e-05, + "lr_decay_style": "cosine", + "lr_decay_iters": null, + "lr_warmup_iters": 0, + "lr_warmup_fraction": 0.05, + "min_lr": 1e-06, + "weight_decay": 0.1, + "clip_grad": 1.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_eps": 1e-08, + "sgd_momentum": 0.9, + "save": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b", + "save_interval": 2000, + "save_retain_interval": null, + "no_save_optim": true, + "no_save_rng": true, + "load": null, + "no_load_optim": false, + "no_load_rng": false, + "finetune": true, + "ckpt_format": "torch_dist", + "no_initialization": true, + "auto_detect_ckpt_format": true, + "exit_on_missing_checkpoint": true, + "async_save": false, + "use_persistent_ckpt_worker": false, + "ckpt_fully_parallel_load": false, + "ckpt_assume_constant_structure": false, + "distributed_backend": "nccl", + "local_rank": 0, + "use_distributed_optimizer": true, + "tensor_model_parallel_size": 4, + "pipeline_model_parallel_size": 1, + "decoder_first_pipeline_num_layers": null, + "decoder_last_pipeline_num_layers": null, + "account_for_embedding_in_pipeline_split": false, + "account_for_loss_in_pipeline_split": false, + "sequence_parallel": true, + "context_parallel_size": 1, + "tp_comm_overlap": false, + "overlap_grad_reduce": false, + "overlap_param_gather": false, + "distributed_timeout_minutes": 300000, + "num_layers_per_virtual_pipeline_stage": null, + "num_virtual_stages_per_pipeline_rank": null, + "microbatch_group_size_per_virtual_pipeline_stage": null, + "pipeline_model_parallel_layout": null, + "num_layers": 48, + "hidden_size": 2048, + "ffn_hidden_size": null, + "num_attention_heads": 32, + "group_query_attention": true, + "num_query_groups": 4, + "softmax_type": "vanilla", + "window_size": null, + "window_attn_skip_freq": null, + "max_position_embeddings": 262144, + "position_embedding_type": "mrope", + "mrope_section": [ + 24, + 20, + 20 + ], + "rotary_base": 5000000, + "rotary_percent": 1.0, + "rotary_interleaved": false, + "normalization": "RMSNorm", + "norm_epsilon": 1e-06, + "swiglu": true, + "quick_geglu": false, + "activation_func_clamp_value": null, + "glu_linear_offset": 0.0, + "untie_embeddings_and_output_weights": true, + "disable_bias_linear": true, + "add_qkv_bias": false, + "attention_dropout": 0.0, + "hidden_dropout": 0.0, + "kv_channels": 128, + "qk_layernorm": true, + "qk_l2_norm": null, + "no_rope_freq": null, + "moe_apply_probs_on_input": null, + "transformer_impl": "transformer_engine", + "num_experts": 128, + "moe_layer_freq": 1, + "moe_ffn_hidden_size": 768, + "moe_shared_expert_intermediate_size": null, + "moe_router_topk": 8, + "moe_router_num_groups": null, + "moe_router_group_topk": null, + "moe_router_pre_softmax": false, + "moe_router_dtype": "fp32", + "moe_router_score_function": "softmax", + "moe_router_bias_update_rate": null, + "moe_router_enable_expert_bias": false, + "moe_router_topk_scaling_factor": null, + "moe_router_load_balancing_type": "aux_loss", + "expert_model_parallel_size": 8, + "expert_tensor_parallel_size": 1, + "moe_token_dispatcher_type": null, + "moe_enable_deepep": false, + "moe_grouped_gemm": true, + "moe_permute_fusion": true, + "moe_aux_loss_coeff": 1e-06, + "moe_z_loss_coeff": null, + "moe_shared_expert_overlap": true, + "moe_layer_recompute": false, + "moe_expert_capacity_factor": 2.0, + "moe_pad_expert_input_to_capacity": false, + "moe_token_drop_policy": null, + "multi_latent_attention": false, + "q_lora_rank": null, + "kv_lora_rank": 32, + "qk_head_dim": 128, + "qk_pos_emb_head_dim": 64, + "mtp_num_layers": null, + "mtp_loss_scaling_factor": 0.1, + "fp8_format": null, + "fp8_recipe": "delayed", + "fp8_amax_history_len": 1024, + "fp8_amax_compute_algo": "max", + "fp8_param_gather": false, + "fp16": false, + "bf16": true, + "apply_query_key_layer_scaling": false, + "attention_softmax_in_fp32": true, + "log_params_norm": false, + "log_throughput": false, + "tensorboard_log_interval": 1, + "tensorboard_queue_size": 50, + "log_timers_to_tensorboard": true, + "no_log_learning_rate_to_tensorboard": false, + "log_validation_ppl_to_tensorboard": true, + "log_memory_to_tensorboard": true, + "logging_level": null, + "wandb_project": "sft_qwen3vl_temporal_grounding", + "wandb_exp_name": "sft_46k_multi_32k_timelens_1epoch_w_cot_30b", + "wandb_save_dir": null, + "eval_iters": -1, + "eval_interval": 2000, + "seq_length": 16384, + "num_workers": 8, + "no_data_sharding": false, + "megatron_extra_kwargs": {}, + "add_version": false, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3-VL-30B-A3B-Instruct", + "model_info": "ModelInfo(model_type='qwen3_moe_vl', model_dir='/opt/tiger/tanyue/highlight_tools/checkpoints/Qwen3-VL-30B-A3B-Instruct', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling={'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default'}, is_moe_model=True, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_moe_vl', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-VL-30B-A3B-Instruct', hf_model_id='Qwen/Qwen3-VL-30B-A3B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-30B-A3B-Thinking', hf_model_id='Qwen/Qwen3-VL-30B-A3B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-30B-A3B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-30B-A3B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-30B-A3B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-30B-A3B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-235B-A22B-Instruct', hf_model_id='Qwen/Qwen3-VL-235B-A22B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-235B-A22B-Thinking', hf_model_id='Qwen/Qwen3-VL-235B-A22B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-235B-A22B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-235B-A22B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-235B-A22B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-235B-A22B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='qwen3_vl', get_function=, model_arch=MultiModelKeys(arch_name='qwen3_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model', 'lm_head'], aligner=['model.visual.merger', 'model.visual.deepstack_merger_list'], vision_tower=['model.visual'], generator=[]), architectures=['Qwen3VLMoeForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, is_reranker=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])", + "model_dir": "/opt/tiger/tanyue/highlight_tools/checkpoints/Qwen3-VL-30B-A3B-Instruct", + "_val_dataset_exists": [], + "hub": "", + "megatron_model_meta": "MegatronModelMeta(megatron_model_type='qwen3_vl', model_types=['qwen3_vl', 'qwen3_moe_vl'], is_multimodal=True, bridge_cls=, model_cls=, get_transformer_layer_spec=None, model_provider=, visual_cls=, get_mtp_block_spec=None, extra_args_provider=None)", + "extra_args": { + "model_dir": "/opt/tiger/tanyue/highlight_tools/checkpoints/Qwen3-VL-30B-A3B-Instruct", + "is_multimodal": true, + "hf_model_type": "qwen3_moe_vl", + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "checkpoints/Qwen3-VL-30B-A3B-Instruct", + "model_type": "qwen3_moe_vl", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": null, + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_type": "default" + }, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_vl", + "system": null, + "max_length": 16384, + "truncation_strategy": "delete", + "max_pixels": null, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_side": "right", + "padding_free": true, + "sequence_parallel_size": 1, + "template_backend": "swift", + "response_prefix": null, + "enable_thinking": null, + "add_non_thinking_prefix": true, + "dataset": [ + "../data/cosmo-cap/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/moment10m/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/vtimellm/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/timelens-100k/train/train_w_cot_reasoning_32k.jsonl" + ], + "val_dataset": [], + "cached_dataset": [], + "cached_val_dataset": [], + "split_dataset_ratio": 0.0, + "data_seed": 42, + "dataset_num_proc": 8, + "load_from_cache_file": false, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "remove_unused_columns": true, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": null, + "temperature": 0.9, + "top_k": 50, + "top_p": 0.9, + "repetition_penalty": 1.0, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "structured_outputs_regex": null, + "ckpt_dir": null, + "lora_modules": [], + "tuner_backend": "peft", + "train_type": "lora", + "adapters": [], + "external_plugins": [], + "model_kwargs": {}, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "custom_register_path": [], + "use_hf": false, + "hub_token": null, + "ddp_timeout": 18000000, + "ddp_backend": null, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": true, + "freeze_parameters": [], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [], + "trainable_parameters_regex": null, + "adapter_load": null, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "use_rslora": false, + "rlhf_type": null, + "ref_load": null, + "ref_adapter_load": null, + "beta": null, + "rpo_alpha": null, + "reference_free": false, + "label_smoothing": 0.0, + "f_divergence_type": "reverse_kl", + "loss_type": null, + "desirable_weight": 1.0, + "undesirable_weight": 1.0, + "calculate_KL": null, + "center_rewards_coefficient": null, + "teacher_model": null, + "teacher_model_type": null, + "teacher_model_revision": null, + "lmbda": 0.5, + "seq_kd": false, + "offload_teacher_model": false, + "sft_alpha": 0.0, + "generation_batch_size": null, + "steps_per_generation": null, + "num_generations": 8, + "num_generations_eval": null, + "max_completion_length": 512, + "importance_sampling_level": "token", + "tau_pos": 1.0, + "tau_neg": 1.05, + "epsilon": 0.2, + "epsilon_high": null, + "delta": null, + "use_vllm": true, + "vllm_mode": null, + "vllm_enable_prefix_caching": true, + "vllm_gpu_memory_utilization": 0.9, + "vllm_tensor_parallel_size": 1, + "vllm_max_model_len": null, + "vllm_enforce_eager": false, + "vllm_limit_mm_per_prompt": null, + "vllm_disable_cascade_attn": false, + "vllm_max_num_seqs": null, + "vllm_mm_processor_cache_gb": null, + "vllm_engine_kwargs": null, + "sleep_level": 0, + "offload_optimizer": false, + "offload_model": false, + "offload_bridge": false, + "vllm_server_base_url": null, + "vllm_server_host": null, + "vllm_server_port": [ + 8000 + ], + "vllm_server_timeout": 240.0, + "vllm_server_group_port": null, + "reward_funcs": [], + "reward_weights": null, + "cosine_min_len_value_wrong": -0.5, + "cosine_max_len_value_wrong": 0.0, + "cosine_min_len_value_correct": 1.0, + "cosine_max_len_value_correct": 0.5, + "cosine_max_len": null, + "repetition_n_grams": 3, + "repetition_max_penalty": -1.0, + "soft_max_length": null, + "soft_cache_length": null, + "dynamic_sample": false, + "max_resample_times": 3, + "overlong_filter": false, + "scale_rewards": "group", + "advantage_estimator": "grpo", + "kl_in_reward": false, + "wandb_log_unique_prompts": null, + "log_completions": false, + "rollout_importance_sampling_mode": null, + "rollout_importance_sampling_threshold": 2.0, + "log_rollout_offpolicy_metrics": false, + "off_policy_sequence_mask_delta": null, + "log_entropy": false, + "top_entropy_quantile": 1.0, + "reward_model": null, + "reward_model_plugin": null, + "sync_ref_model": false, + "ref_model_sync_steps": 512, + "ref_model_mixup_alpha": 0.6, + "async_generate": false, + "move_model_batches": null, + "multi_turn_scheduler": null, + "max_turns": null, + "completion_length_limit_scope": "per_round", + "vllm_server_pass_dataset": false, + "num_iterations": 1, + "check_model": true, + "padded_vocab_size": 151936, + "initialize_embedding": false, + "mlp_padding_free": false, + "load_safetensors": true, + "save_safetensors": true, + "ref_model": null, + "ref_adapters": [], + "merge_lora": true, + "max_shard_size": "5GB", + "train_dataloader_shuffle": true, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": true, + "dataloader_prefetch_factor": 2, + "group_by_length": false, + "architectures": "Qwen3VLMoeForConditionalGeneration", + "llm_architectures": "Qwen3VLMoeForConditionalGeneration", + "max_epochs": 1, + "enable_dft_loss": false, + "enable_channel_loss": false, + "save_strategy": "steps", + "original_max_position_embeddings": null, + "partial_rotary_factor": null, + "use_shared_expert_gate": false, + "report_to": "wandb", + "vit_gradient_checkpointing": true, + "vit_lr": null, + "aligner_lr": null, + "gradient_checkpointing_kwargs": null, + "linear_num_value_heads": null, + "linear_num_key_heads": null, + "linear_key_head_dim": null, + "linear_value_head_dim": null, + "linear_conv_kernel_dim": null, + "layer_types": null, + "apply_wd_to_qk_layernorm": false, + "mrope_interleaved": true, + "add_version": false + } +} \ No newline at end of file diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772166638.n124-174-062.999503.0 b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772166638.n124-174-062.999503.0 new file mode 100644 index 0000000000000000000000000000000000000000..cdd8f35e0b6ab30d18880934a24bf565c1ad1e80 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772166638.n124-174-062.999503.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4e1c57bb317639935a51362d5bf870bfeb13ae05f61012a7bab113192ce149bd +size 88 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772172078.n124-174-062.1051311.0 b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772172078.n124-174-062.1051311.0 new file mode 100644 index 0000000000000000000000000000000000000000..6b0f06b8818a7cbbad9a39712c2c4f12026a9bab --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772172078.n124-174-062.1051311.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:61f19d0ebc13643c9c569887e77e6d584efefd95f1f2d58416613da212b97c92 +size 73481 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772172728.n124-174-062.1098154.0 b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772172728.n124-174-062.1098154.0 new file mode 100644 index 0000000000000000000000000000000000000000..ff8ae19ceba1bc7929c38a9da424bacbece36bfa --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772172728.n124-174-062.1098154.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f1710479446f93086be4fdecd8dfe4cb5a7851d9f4f30482ab2f6e1bce86072e +size 73484 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772173357.n124-174-062.1161238.0 b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772173357.n124-174-062.1161238.0 new file mode 100644 index 0000000000000000000000000000000000000000..9183c5028168ac942427f3398bc63dc835ed02f1 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772173357.n124-174-062.1161238.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7adcd66c8d677ab21aab0b61881fdb4674fdd2618675f3fe15aa5975aaa11e11 +size 73484 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772174808.n124-174-062.1206311.0 b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772174808.n124-174-062.1206311.0 new file mode 100644 index 0000000000000000000000000000000000000000..730bed5ef680fdc3a64589abcd41773a020a2163 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772174808.n124-174-062.1206311.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9a174ca7d1dcbdf8d43f05ce0d58b977809059a5da2f1fb8adaa03739e46137f +size 73484 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772175303.n124-174-062.1237489.0 b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772175303.n124-174-062.1237489.0 new file mode 100644 index 0000000000000000000000000000000000000000..4ba69f4d35062788bca43d896d6dc03ac874a397 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772175303.n124-174-062.1237489.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b1045ea3a963e47160948c43b14c14471c8ca6dc4290c8b66e891390828fd659 +size 88 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772175960.n124-174-062.1248914.0 b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772175960.n124-174-062.1248914.0 new file mode 100644 index 0000000000000000000000000000000000000000..3351e0beb2e36bdfeacd58bdd4a81dde4271bf67 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/runs/events.out.tfevents.1772175960.n124-174-062.1248914.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:934da4579d8f5405d349999a977bfe00335cef443f240d3f13ba123dff08f8a1 +size 73484 diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/v1-20260223-030938/args.json b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/v1-20260223-030938/args.json new file mode 100644 index 0000000000000000000000000000000000000000..b8da5ba849386d240ffebe1b771780b76454557b --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/v1-20260223-030938/args.json @@ -0,0 +1,398 @@ +{ + "output_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/v1-20260223-030938", + "overwrite_output_dir": false, + "do_train": false, + "do_eval": false, + "do_predict": false, + "eval_strategy": "no", + "prediction_loss_only": false, + "per_device_train_batch_size": 4, + "per_device_eval_batch_size": 4, + "per_gpu_train_batch_size": null, + "per_gpu_eval_batch_size": null, + "gradient_accumulation_steps": 4, + "eval_accumulation_steps": null, + "eval_delay": 0, + "torch_empty_cache_steps": null, + "learning_rate": 1e-05, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "max_grad_norm": 1.0, + "num_train_epochs": 1.0, + "max_steps": -1, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_ratio": 0.03, + "warmup_steps": 0, + "log_level": "passive", + "log_level_replica": "warning", + "log_on_each_node": true, + "logging_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/v1-20260223-030938/runs", + "logging_strategy": "steps", + "logging_first_step": true, + "logging_steps": 5, + "logging_nan_inf_filter": true, + "save_strategy": "epoch", + "save_steps": 500, + "save_total_limit": 1, + "save_safetensors": true, + "save_on_each_node": false, + "save_only_model": false, + "restore_callback_states_from_checkpoint": false, + "no_cuda": false, + "use_cpu": false, + "use_mps_device": false, + "seed": 42, + "data_seed": 42, + "jit_mode_eval": false, + "bf16": true, + "fp16": false, + "fp16_opt_level": "O1", + "half_precision_backend": "auto", + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": null, + "local_rank": 0, + "ddp_backend": null, + "tpu_num_cores": null, + "tpu_metrics_debug": false, + "debug": null, + "dataloader_drop_last": false, + "eval_steps": null, + "dataloader_num_workers": 24, + "dataloader_prefetch_factor": 2, + "past_index": -1, + "run_name": "sft_46k_multi_32k_timelens_1epoch_w_cot_30b", + "disable_tqdm": null, + "remove_unused_columns": true, + "label_names": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "fsdp": [], + "fsdp_min_num_params": 0, + "fsdp_config": null, + "fsdp_transformer_layer_cls_to_wrap": null, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "deepspeed": { + "fp16": { + "enabled": "auto", + "loss_scale": 0, + "loss_scale_window": 1000, + "initial_scale_power": 16, + "hysteresis": 2, + "min_loss_scale": 1 + }, + "bf16": { + "enabled": "auto" + }, + "zero_optimization": { + "stage": 3, + "offload_optimizer": { + "device": "none", + "pin_memory": true + }, + "offload_param": { + "device": "none", + "pin_memory": true + }, + "overlap_comm": false, + "contiguous_gradients": true, + "sub_group_size": 1000000000.0, + "reduce_bucket_size": "auto", + "zero_quantized_weights": false, + "zero_quantized_gradients": false, + "stage3_prefetch_bucket_size": "auto", + "stage3_param_persistence_threshold": "auto", + "stage3_max_live_parameters": 1000000000.0, + "stage3_max_reuse_distance": 1000000000.0, + "stage3_gather_16bit_weights_on_model_save": true + }, + "gradient_accumulation_steps": "auto", + "gradient_clipping": "auto", + "steps_per_print": 2000, + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "wall_clock_breakdown": false + }, + "label_smoothing_factor": 0.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "adafactor": false, + "group_by_length": false, + "length_column_name": "length", + "report_to": [ + "wandb" + ], + "project": "huggingface", + "trackio_space_id": "trackio", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": false, + "skip_memory_metrics": true, + "use_legacy_prediction_loop": false, + "push_to_hub": false, + "resume_from_checkpoint": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_token": null, + "hub_private_repo": null, + "hub_always_push": false, + "hub_revision": null, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "include_inputs_for_metrics": false, + "include_for_metrics": [], + "eval_do_concat_batches": true, + "fp16_backend": "auto", + "push_to_hub_model_id": null, + "push_to_hub_organization": null, + "push_to_hub_token": null, + "mp_parameters": "", + "auto_find_batch_size": false, + "full_determinism": false, + "torchdynamo": null, + "ray_scope": "last", + "ddp_timeout": 18000000, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "include_tokens_per_second": false, + "include_num_input_tokens_seen": false, + "neftune_noise_alpha": null, + "optim_target_modules": null, + "batch_eval_metrics": false, + "eval_on_start": false, + "use_liger_kernel": false, + "liger_kernel_config": null, + "eval_use_gather_object": false, + "average_tokens_across_devices": true, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": null, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "metric": [ + "temporal_grounding" + ], + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "/mnt/bn/strategy-mllm-train/user/zjn/checkpoints/Qwen3-VL-30B-A3B-Instruct", + "model_type": "qwen3_moe_vl", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": "flash_attn", + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_vl", + "system": null, + "max_length": 262144, + "truncation_strategy": "delete", + "max_pixels": null, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_free": true, + "padding_side": "right", + "loss_scale": "default", + "sequence_parallel_size": 8, + "response_prefix": null, + "template_backend": "swift", + "dataset": [ + "../data/cosmo-cap/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/moment10m/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/vtimellm/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "../data/timelens-100k/train/train_w_cot_reasoning_32k.jsonl" + ], + "val_dataset": [], + "cached_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 1, + "load_from_cache_file": true, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "ckpt_dir": null, + "lora_modules": [], + "train_type": "full", + "adapters": [], + "external_plugins": [], + "model_kwargs": {}, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "custom_register_path": [], + "use_hf": true, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [ + "model.visual.merger", + "model.visual.deepstack_merger_list" + ], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": false, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": null, + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_lark_webhook_url": null, + "swanlab_lark_secret": null, + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "early_stop_interval": null, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3-VL-30B-A3B-Instruct", + "model_info": "ModelInfo(model_type='qwen3_moe_vl', model_dir='/mnt/bn/strategy-mllm-train/user/zjn/checkpoints/Qwen3-VL-30B-A3B-Instruct', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling={'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default'}, is_moe_model=True, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_moe_vl', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-VL-30B-A3B-Instruct', hf_model_id='Qwen/Qwen3-VL-30B-A3B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-30B-A3B-Thinking', hf_model_id='Qwen/Qwen3-VL-30B-A3B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-30B-A3B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-30B-A3B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-30B-A3B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-30B-A3B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-235B-A22B-Instruct', hf_model_id='Qwen/Qwen3-VL-235B-A22B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-235B-A22B-Thinking', hf_model_id='Qwen/Qwen3-VL-235B-A22B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-235B-A22B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-235B-A22B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-235B-A22B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-235B-A22B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='qwen3_vl', get_function=, model_arch=MultiModelKeys(arch_name='qwen3_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model'], aligner=['model.visual.merger', 'model.visual.deepstack_merger_list'], vision_tower=['model.visual'], generator=[]), architectures=['Qwen3VLMoeForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, is_reranker=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])", + "model_dir": "/mnt/bn/strategy-mllm-train/user/zjn/checkpoints/Qwen3-VL-30B-A3B-Instruct", + "_val_dataset_exists": [], + "hub": "", + "evaluation_strategy": "epoch", + "training_args": "Seq2SeqTrainingArguments(output_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/v1-20260223-030938', overwrite_output_dir=False, do_train=False, do_eval=False, do_predict=False, eval_strategy=, prediction_loss_only=False, per_device_train_batch_size=4, per_device_eval_batch_size=4, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=4, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=1e-05, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=1.0, max_steps=-1, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_ratio=0.03, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/v1-20260223-030938/runs', logging_strategy=, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=, save_steps=500, save_total_limit=1, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, jit_mode_eval=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=None, dataloader_num_workers=24, dataloader_prefetch_factor=2, past_index=-1, run_name='sft_46k_multi_32k_timelens_1epoch_w_cot_30b', disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, deepspeed={'fp16': {'enabled': 'auto', 'loss_scale': 0, 'loss_scale_window': 1000, 'initial_scale_power': 16, 'hysteresis': 2, 'min_loss_scale': 1}, 'bf16': {'enabled': 'auto'}, 'zero_optimization': {'stage': 3, 'offload_optimizer': {'device': 'none', 'pin_memory': True}, 'offload_param': {'device': 'none', 'pin_memory': True}, 'overlap_comm': False, 'contiguous_gradients': True, 'sub_group_size': 1000000000.0, 'reduce_bucket_size': 'auto', 'zero_quantized_weights': False, 'zero_quantized_gradients': False, 'stage3_prefetch_bucket_size': 'auto', 'stage3_param_persistence_threshold': 'auto', 'stage3_max_live_parameters': 1000000000.0, 'stage3_max_reuse_distance': 1000000000.0, 'stage3_gather_16bit_weights_on_model_save': True}, 'gradient_accumulation_steps': 'auto', 'gradient_clipping': 'auto', 'steps_per_print': 2000, 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'wall_clock_breakdown': False}, label_smoothing_factor=0.0, optim=, optim_args=None, adafactor=False, group_by_length=False, length_column_name='length', report_to=['wandb'], project='huggingface', trackio_space_id='trackio', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint=None, hub_model_id=None, hub_strategy=, hub_token=None, hub_private_repo=None, hub_always_push=False, hub_revision=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=18000000, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, include_tokens_per_second=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, liger_kernel_config=None, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=True, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, check_model=True, acc_strategy='token', train_dataloader_shuffle=True, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, metric=['temporal_grounding'], eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, use_flash_ckpt=False, sft_alpha=0, chord_sft_dataset=[], chord_sft_per_device_train_batch_size=None, chord_enable_phi_function=False, chord_mu_warmup_steps=None, chord_mu_decay_steps=None, chord_mu_peak=None, chord_mu_valley=None, train_type='full', local_repo_path=None, galore_config=None, padding_side='right', padding_free=True, task_type='causal_lm', problem_type=None)" +} \ No newline at end of file diff --git a/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/v1-20260223-030938/logging.jsonl b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/v1-20260223-030938/logging.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..4a30e4397b1fd572060ffce8b76629d335529201 --- /dev/null +++ b/posttrain_models/sft_46k_multi_32k_timelens_1epoch_w_cot_30b/v1-20260223-030938/logging.jsonl @@ -0,0 +1 @@ +{"model_parameter_info": "Qwen3VLMoeForConditionalGeneration: 31070.7540M Params (30654.8626M Trainable [98.6615%]), 0.0001M Buffers.", "last_model_checkpoint": null, "best_model_checkpoint": null, "best_metric": null, "global_step": 0, "log_history": [], "memory": null} diff --git a/posttrain_models/sft_qwen3omni_30b_a3b_omtg_wcot_1epoch/args.json b/posttrain_models/sft_qwen3omni_30b_a3b_omtg_wcot_1epoch/args.json new file mode 100644 index 0000000000000000000000000000000000000000..1326742744d4d38c9a4e332e67d16b4e955c1943 --- /dev/null +++ b/posttrain_models/sft_qwen3omni_30b_a3b_omtg_wcot_1epoch/args.json @@ -0,0 +1,392 @@ +{ + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "/mnt/bn/strategy-mllm-train/user/tanyue/models/pretrained_models/Qwen3-Omni-30B-A3B-Instruct", + "model_type": "qwen3_omni_moe", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": null, + "experts_impl": null, + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_omni", + "system": null, + "max_length": 65536, + "truncation_strategy": "delete", + "max_pixels": null, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_side": "right", + "padding_free": true, + "loss_scale": "default", + "sequence_parallel_size": 1, + "template_backend": "swift", + "response_prefix": null, + "enable_thinking": null, + "add_non_thinking_prefix": true, + "dataset": [ + "data/cosmo-cap/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "data/moment10m/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "data/vtimellm/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "data/timelens-100k/train/train_w_cot_reasoning_32k.jsonl" + ], + "val_dataset": [], + "cached_dataset": [], + "cached_val_dataset": [], + "split_dataset_ratio": 0.0, + "data_seed": 42, + "dataset_num_proc": 8, + "load_from_cache_file": false, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "remove_unused_columns": true, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": null, + "temperature": 0.9, + "top_k": 50, + "top_p": 0.9, + "repetition_penalty": 1.0, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "structured_outputs_regex": null, + "tuner_backend": "peft", + "tuner_type": "lora", + "train_type": "lora", + "adapters": [], + "external_plugins": [], + "custom_register_path": [], + "seed": 42, + "model_kwargs": {}, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "use_hf": false, + "hub_token": null, + "ddp_timeout": 18000000, + "ddp_backend": "nccl", + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": true, + "freeze_parameters": [], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [], + "trainable_parameters_regex": null, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "use_rslora": false, + "rlhf_type": null, + "loss_type": null, + "mcore_ref_model": null, + "mcore_ref_adapter": null, + "beta": null, + "rpo_alpha": null, + "reference_free": false, + "label_smoothing": 0.0, + "f_divergence_type": "reverse_kl", + "desirable_weight": 1.0, + "undesirable_weight": 1.0, + "calculate_KL": null, + "center_rewards_coefficient": null, + "teacher_model": null, + "teacher_model_type": null, + "teacher_model_revision": null, + "teacher_model_server": null, + "gkd_logits_topk": null, + "lmbda": 0.5, + "seq_kd": false, + "offload_teacher_model": false, + "sft_alpha": 0.0, + "generation_batch_size": null, + "steps_per_generation": null, + "num_generations": 8, + "num_generations_eval": null, + "max_completion_length": 512, + "importance_sampling_level": "token", + "tau_pos": 1.0, + "tau_neg": 1.05, + "epsilon": 0.2, + "epsilon_high": null, + "delta": null, + "use_vllm": true, + "vllm_mode": null, + "vllm_enable_prefix_caching": true, + "vllm_gpu_memory_utilization": 0.9, + "vllm_tensor_parallel_size": 1, + "vllm_max_model_len": null, + "vllm_enforce_eager": false, + "vllm_limit_mm_per_prompt": null, + "vllm_disable_cascade_attn": false, + "vllm_max_num_seqs": null, + "vllm_mm_processor_cache_gb": null, + "vllm_engine_kwargs": null, + "sleep_level": 0, + "offload_optimizer": false, + "offload_model": false, + "offload_bridge": false, + "vllm_server_base_url": null, + "vllm_server_host": null, + "vllm_server_port": [ + 8000 + ], + "vllm_server_timeout": 240.0, + "vllm_server_group_port": null, + "reward_funcs": [], + "reward_weights": null, + "cosine_min_len_value_wrong": -0.5, + "cosine_max_len_value_wrong": 0.0, + "cosine_min_len_value_correct": 1.0, + "cosine_max_len_value_correct": 0.5, + "cosine_max_len": null, + "repetition_n_grams": 3, + "repetition_max_penalty": -1.0, + "soft_max_length": null, + "soft_cache_length": null, + "dynamic_sample": false, + "max_resample_times": 3, + "overlong_filter": false, + "scale_rewards": "group", + "advantage_estimator": "grpo", + "kl_in_reward": false, + "wandb_log_unique_prompts": null, + "log_completions": false, + "rollout_importance_sampling_mode": null, + "rollout_importance_sampling_threshold": 2.0, + "log_rollout_offpolicy_metrics": false, + "off_policy_sequence_mask_delta": null, + "log_entropy": false, + "top_entropy_quantile": 1.0, + "reward_model": null, + "reward_model_plugin": null, + "sync_ref_model": false, + "ref_model_sync_steps": 512, + "ref_model_mixup_alpha": 0.6, + "async_generate": false, + "move_model_batches": null, + "multi_turn_scheduler": null, + "max_turns": null, + "completion_length_limit_scope": "per_round", + "vllm_server_pass_dataset": false, + "num_iterations": 1, + "micro_batch_size": 1, + "global_batch_size": 8, + "recompute_granularity": "full", + "recompute_method": "uniform", + "recompute_num_layers": 1, + "recompute_modules": [ + "core_attn" + ], + "train_iters": null, + "num_train_epochs": 4, + "masked_softmax_fusion": true, + "bias_dropout_fusion": true, + "bias_activation_fusion": true, + "apply_rope_fusion": false, + "gradient_accumulation_fusion": false, + "cross_entropy_loss_fusion": true, + "cross_entropy_fusion_impl": "te", + "calculate_per_token_loss": true, + "attention_backend": "", + "optimizer": "adam", + "optimizer_cpu_offload": false, + "optimizer_offload_fraction": 1.0, + "use_precision_aware_optimizer": false, + "main_grads_dtype": "float32", + "main_params_dtype": "float32", + "exp_avg_dtype": "float32", + "exp_avg_sq_dtype": "float32", + "manual_gc": false, + "manual_gc_steps": 0, + "manual_gc_eval": true, + "train_dataloader_shuffle": true, + "dataloader_num_workers": 8, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": true, + "dataloader_prefetch_factor": 2, + "data_sharding": false, + "group_by_length": false, + "te_rng_tracker": false, + "data_parallel_random_init": false, + "mlp_padding_free": false, + "lr_warmup_init": 0.0, + "lr": 0.0001, + "lr_decay_style": "cosine", + "lr_decay_iters": null, + "lr_warmup_iters": 0, + "lr_warmup_fraction": 0.05, + "min_lr": 1e-05, + "lr_wsd_decay_style": "exponential", + "lr_wsd_decay_iters": null, + "weight_decay": 0.1, + "weight_decay_incr_style": "constant", + "start_weight_decay": 0.1, + "end_weight_decay": 0.1, + "clip_grad": 1.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_eps": 1e-08, + "sgd_momentum": 0.9, + "output_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3omni_30b_a3b_omtg_wcot_1epoch", + "save_steps": 1000, + "no_save_optim": true, + "no_save_rng": true, + "mcore_model": null, + "mcore_adapter": null, + "no_load_optim": false, + "no_load_rng": false, + "finetune": true, + "perform_initialization": false, + "use_cpu_initialization": false, + "async_save": false, + "save_total_limit": null, + "metric_for_best_model": "loss", + "greater_is_better": false, + "use_persistent_ckpt_worker": false, + "dist_ckpt_save_pre_mcore_014": false, + "dist_ckpt_optim_fully_reshardable": false, + "distrib_optim_fully_reshardable_mem_efficient": false, + "local_rank": 0, + "use_distributed_optimizer": true, + "tensor_model_parallel_size": 4, + "pipeline_model_parallel_size": 2, + "decoder_first_pipeline_num_layers": null, + "decoder_last_pipeline_num_layers": null, + "account_for_embedding_in_pipeline_split": false, + "account_for_loss_in_pipeline_split": false, + "overlap_p2p_comm": false, + "align_param_gather": false, + "sequence_parallel": true, + "context_parallel_size": 1, + "tp_comm_overlap": false, + "overlap_grad_reduce": false, + "overlap_param_gather": false, + "overlap_param_gather_with_optimizer_step": false, + "align_grad_reduce": true, + "virtual_pipeline_model_parallel_size": null, + "microbatch_group_size_per_vp_stage": null, + "pipeline_model_parallel_layout": null, + "expert_model_parallel_size": 2, + "expert_tensor_parallel_size": 1, + "report_to": [ + "wandb" + ], + "logging_steps": 5, + "tensorboard_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3omni_30b_a3b_omtg_wcot_1epoch/runs", + "tensorboard_queue_size": 50, + "wandb_project": "highlight_qwen3omni", + "wandb_exp_name": "sft_qwen3omni_30b_a3b_omtg_wcot_1epoch", + "swanlab_project": "megatron-swift", + "swanlab_exp_name": null, + "eval_iters": -1, + "eval_steps": 100000, + "fp8_format": null, + "fp8_recipe": "delayed", + "fp8_amax_history_len": 1024, + "fp8_amax_compute_algo": "max", + "fp8_param_gather": false, + "fp16": false, + "bf16": true, + "apply_query_key_layer_scaling": false, + "attention_softmax_in_fp32": true, + "accumulate_allreduce_grads_in_fp32": true, + "moe_router_load_balancing_type": null, + "moe_router_dtype": "fp32", + "moe_token_dispatcher_type": null, + "moe_enable_deepep": false, + "moe_grouped_gemm": true, + "moe_permute_fusion": true, + "moe_aux_loss_coeff": 1e-06, + "moe_z_loss_coeff": null, + "moe_shared_expert_overlap": true, + "moe_layer_recompute": false, + "moe_expert_capacity_factor": null, + "moe_pad_expert_input_to_capacity": false, + "moe_token_drop_policy": null, + "mtp_num_layers": null, + "mtp_loss_scaling_factor": 0.1, + "save_safetensors": true, + "ref_model": null, + "ref_adapters": [], + "merge_lora": true, + "max_shard_size": "5GB", + "vit_gradient_checkpointing": false, + "vit_lr": null, + "aligner_lr": null, + "gradient_checkpointing_kwargs": null, + "dsa_indexer_loss_coeff": null, + "dsa_indexer_use_sparse_loss": false, + "check_model": true, + "apply_wd_to_qk_layernorm": false, + "enable_dft_loss": false, + "enable_channel_loss": false, + "save_strategy": "steps", + "callbacks": [ + "print", + "default_flow", + "wandb" + ], + "add_version": false, + "create_checkpoint_symlink": false, + "swift_version": "4.1.0.dev0", + "ckpt_dir": null, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3-Omni-30B-A3B-Instruct", + "model_info": "ModelInfo(model_type='qwen3_omni_moe', model_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/pretrained_models/Qwen3-Omni-30B-A3B-Instruct', torch_dtype=torch.bfloat16, max_model_len=65536, quant_method=None, quant_bits=None, rope_scaling={'interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default', 'type': 'default'}, is_moe_model=True, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_omni_moe', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-Omni-30B-A3B-Instruct', hf_model_id='Qwen/Qwen3-Omni-30B-A3B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-Omni-30B-A3B-Thinking', hf_model_id='Qwen/Qwen3-Omni-30B-A3B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-Omni-30B-A3B-Captioner', hf_model_id='Qwen/Qwen3-Omni-30B-A3B-Captioner', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_omni', ignore_patterns=None, requires=None, tags=[])], loader=, template='qwen3_omni', model_arch=MultiModelKeys(arch_name='qwen3_omni', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['thinker.model', 'thinker.lm_head'], aligner=['thinker.audio_tower.proj1', 'thinker.audio_tower.proj2', 'thinker.visual.merger', 'thinker.visual.merger_list'], vision_tower=['thinker.audio_tower', 'thinker.visual'], generator=['talker', 'code2wav']), architectures=['Qwen3OmniMoeForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57.dev0', 'soundfile', 'decord', 'qwen_omni_utils>=0.0.9'], tags=[])", + "model_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/pretrained_models/Qwen3-Omni-30B-A3B-Instruct", + "template_meta": "QwenTemplateMeta(template_type='qwen3_omni', prefix=[], prompt=['<|im_start|>user\\n{{QUERY}}<|im_end|>\\n<|im_start|>assistant\\n'], chat_sep=['<|im_end|>\\n'], suffix=['<|im_end|>\\n'], template_cls=, system_prefix=['<|im_start|>system\\n{{SYSTEM}}<|im_end|>\\n'], default_system=None, auto_add_bos=False, stop_words=['<|endoftext|>'], agent_template='hermes', is_thinking=False, thinking_prefix='\\n', non_thinking_prefix='', history_thinking_prefix='')", + "_val_dataset_exists": false, + "hub": "", + "seq_length": 65536, + "is_multimodal": true, + "megatron_model_meta": "MegatronModelMeta(megatron_model_type='qwen3_omni', model_types=['qwen3_omni_moe'], bridge_cls=, visual_cls=, is_multimodal=True, loader=)", + "fp8": null, + "params_dtype": "bfloat16", + "world_size": 8, + "data_parallel_size": 1, + "num_microbatches": 8 +} \ No newline at end of file diff --git a/posttrain_models/sft_qwen3omni_30b_a3b_omtg_wcot_1epoch/logging.jsonl b/posttrain_models/sft_qwen3omni_30b_a3b_omtg_wcot_1epoch/logging.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..24f7c7d82e1e30f845e849a8b30c1a1f2d3129b1 --- /dev/null +++ b/posttrain_models/sft_qwen3omni_30b_a3b_omtg_wcot_1epoch/logging.jsonl @@ -0,0 +1,39 @@ +{"loss": 0.79752916, "grad_norm": 0.71861428, "learning_rate": 5e-08, "load_balancing_loss": 1.16632819, "iteration": "1/39476", "elapsed_time": "3m 39s", "remaining_time": "99d 22h 14m 25s", "memory(GiB)": 34.3, "train_speed(s/it)": 218.712225} +{"loss": 0.7969541, "grad_norm": 0.7025395, "learning_rate": 5e-08, "load_balancing_loss": 1.17176342, "iteration": "1/39476", "elapsed_time": "2m 53s", "remaining_time": "79d 2h 43m 50s", "memory(GiB)": 29.87, "train_speed(s/it)": 173.158455} +{"loss": 0.62428027, "grad_norm": 0.46316177, "learning_rate": 2.5e-07, "load_balancing_loss": 1.15305471, "iteration": "5/39476", "elapsed_time": "7m 48s", "remaining_time": "42d 17h 59m 11s", "memory(GiB)": 38.06, "train_speed(s/it)": 93.576316} +{"loss": 0.61962575, "grad_norm": 0.46594495, "learning_rate": 5.1e-07, "load_balancing_loss": 1.15364075, "iteration": "10/39476", "elapsed_time": "13m 8s", "remaining_time": "36d 0h 5m 40s", "memory(GiB)": 38.53, "train_speed(s/it)": 78.820746} +{"loss": 0.64404851, "grad_norm": 0.73658067, "learning_rate": 7.6e-07, "load_balancing_loss": 1.14122546, "iteration": "15/39476", "elapsed_time": "18m 20s", "remaining_time": "33d 12h 11m 42s", "memory(GiB)": 38.53, "train_speed(s/it)": 73.366171} +{"loss": 0.70058191, "grad_norm": 0.47292313, "learning_rate": 1.01e-06, "load_balancing_loss": 1.14928579, "iteration": "20/39476", "elapsed_time": "23m 4s", "remaining_time": "31d 14h 40m 53s", "memory(GiB)": 38.53, "train_speed(s/it)": 69.222758} +{"loss": 0.68722153, "grad_norm": 0.45446816, "learning_rate": 1.27e-06, "load_balancing_loss": 1.14899564, "iteration": "25/39476", "elapsed_time": "28m 23s", "remaining_time": "31d 2h 32m 32s", "memory(GiB)": 38.53, "train_speed(s/it)": 68.123794} +{"loss": 0.68069631, "grad_norm": 0.59046787, "learning_rate": 1.52e-06, "load_balancing_loss": 1.15396547, "iteration": "30/39476", "elapsed_time": "33m 25s", "remaining_time": "30d 12h 10m 34s", "memory(GiB)": 38.53, "train_speed(s/it)": 66.821316} +{"loss": 0.74488693, "grad_norm": 0.75896949, "learning_rate": 1.77e-06, "load_balancing_loss": 1.15226734, "iteration": "35/39476", "elapsed_time": "37m 42s", "remaining_time": "29d 12h 2m 52s", "memory(GiB)": 38.53, "train_speed(s/it)": 64.627466} +{"loss": 0.59542578, "grad_norm": 0.53146434, "learning_rate": 2.03e-06, "load_balancing_loss": 1.14890432, "iteration": "40/39476", "elapsed_time": "42m 48s", "remaining_time": "29d 7h 9m 12s", "memory(GiB)": 38.53, "train_speed(s/it)": 64.188864} +{"loss": 0.56059027, "grad_norm": 0.50748771, "learning_rate": 2.28e-06, "load_balancing_loss": 1.16164958, "iteration": "45/39476", "elapsed_time": "47m 40s", "remaining_time": "29d 0h 4m 26s", "memory(GiB)": 38.53, "train_speed(s/it)": 63.550651} +{"loss": 0.60602075, "grad_norm": 0.48416972, "learning_rate": 2.53e-06, "load_balancing_loss": 1.16343558, "iteration": "50/39476", "elapsed_time": "53m 16s", "remaining_time": "29d 3h 55m 20s", "memory(GiB)": 38.53, "train_speed(s/it)": 63.910109} +{"loss": 0.67060959, "grad_norm": 0.57610112, "learning_rate": 2.79e-06, "load_balancing_loss": 1.14079094, "iteration": "55/39476", "elapsed_time": "57m 39s", "remaining_time": "28d 16h 35m 2s", "memory(GiB)": 38.53, "train_speed(s/it)": 62.882771} +{"loss": 0.68030119, "grad_norm": 0.7307722, "learning_rate": 3.04e-06, "load_balancing_loss": 1.14573693, "iteration": "60/39476", "elapsed_time": "1h 2m 12s", "remaining_time": "28d 9h 3m 47s", "memory(GiB)": 38.53, "train_speed(s/it)": 62.203849} +{"loss": 0.63000673, "grad_norm": 0.49322408, "learning_rate": 3.29e-06, "load_balancing_loss": 1.16319513, "iteration": "65/39476", "elapsed_time": "1h 7m 0s", "remaining_time": "28d 5h 3m 40s", "memory(GiB)": 38.53, "train_speed(s/it)": 61.846179} +{"loss": 0.65486592, "grad_norm": 0.53418535, "learning_rate": 3.55e-06, "load_balancing_loss": 1.15429544, "iteration": "70/39476", "elapsed_time": "1h 12m 6s", "remaining_time": "28d 4h 28m 5s", "memory(GiB)": 38.53, "train_speed(s/it)": 61.799859} +{"loss": 0.58107656, "grad_norm": 0.54243851, "learning_rate": 3.8e-06, "load_balancing_loss": 1.15346587, "iteration": "75/39476", "elapsed_time": "1h 16m 33s", "remaining_time": "27d 22h 17m 51s", "memory(GiB)": 38.53, "train_speed(s/it)": 61.243893} +{"loss": 0.61606735, "grad_norm": 0.54644316, "learning_rate": 4.05e-06, "load_balancing_loss": 1.1544584, "iteration": "80/39476", "elapsed_time": "1h 21m 16s", "remaining_time": "27d 18h 59m 25s", "memory(GiB)": 38.53, "train_speed(s/it)": 60.949467} +{"loss": 0.55808055, "grad_norm": 0.58085799, "learning_rate": 4.31e-06, "load_balancing_loss": 1.15309417, "iteration": "85/39476", "elapsed_time": "1h 25m 60s", "remaining_time": "27d 16h 12m 57s", "memory(GiB)": 38.53, "train_speed(s/it)": 60.703628} +{"loss": 0.64460921, "grad_norm": 0.53109646, "learning_rate": 4.56e-06, "load_balancing_loss": 1.14504063, "iteration": "90/39476", "elapsed_time": "1h 30m 22s", "remaining_time": "27d 11h 5m 52s", "memory(GiB)": 38.53, "train_speed(s/it)": 60.243534} +{"loss": 0.560381, "grad_norm": 0.55679244, "learning_rate": 4.81e-06, "load_balancing_loss": 1.13832664, "iteration": "95/39476", "elapsed_time": "1h 35m 1s", "remaining_time": "27d 8h 25m 57s", "memory(GiB)": 38.53, "train_speed(s/it)": 60.00755} +{"loss": 0.50794268, "grad_norm": 0.45426488, "learning_rate": 5.07e-06, "load_balancing_loss": 1.13994551, "iteration": "100/39476", "elapsed_time": "1h 39m 34s", "remaining_time": "27d 5h 24m 51s", "memory(GiB)": 38.53, "train_speed(s/it)": 59.739195} +{"loss": 0.55100721, "grad_norm": 0.36566952, "learning_rate": 5.32e-06, "load_balancing_loss": 1.14436269, "iteration": "105/39476", "elapsed_time": "1h 44m 12s", "remaining_time": "27d 3h 10m 59s", "memory(GiB)": 38.53, "train_speed(s/it)": 59.542784} +{"loss": 0.45714629, "grad_norm": 0.5526337, "learning_rate": 5.57e-06, "load_balancing_loss": 1.15952468, "iteration": "110/39476", "elapsed_time": "1h 48m 42s", "remaining_time": "27d 0h 22m 51s", "memory(GiB)": 38.53, "train_speed(s/it)": 59.294097} +{"loss": 0.47067213, "grad_norm": 0.58670747, "learning_rate": 5.83e-06, "load_balancing_loss": 1.15042448, "iteration": "115/39476", "elapsed_time": "1h 53m 45s", "remaining_time": "27d 0h 53m 50s", "memory(GiB)": 38.53, "train_speed(s/it)": 59.348848} +{"loss": 0.40844089, "grad_norm": 0.46943074, "learning_rate": 6.08e-06, "load_balancing_loss": 1.15037131, "iteration": "120/39476", "elapsed_time": "1h 58m 45s", "remaining_time": "27d 1h 3m 38s", "memory(GiB)": 38.53, "train_speed(s/it)": 59.371329} +{"loss": 0.40966654, "grad_norm": 0.49872005, "learning_rate": 6.33e-06, "load_balancing_loss": 1.15591764, "iteration": "125/39476", "elapsed_time": "2h 3m 49s", "remaining_time": "27d 1h 37m 2s", "memory(GiB)": 38.53, "train_speed(s/it)": 59.429797} +{"loss": 0.40699118, "grad_norm": 0.36618179, "learning_rate": 6.59e-06, "load_balancing_loss": 1.15782666, "iteration": "130/39476", "elapsed_time": "2h 9m 17s", "remaining_time": "27d 4h 10m 55s", "memory(GiB)": 38.53, "train_speed(s/it)": 59.672021} +{"loss": 0.39758924, "grad_norm": 0.36957055, "learning_rate": 6.84e-06, "load_balancing_loss": 1.14350808, "iteration": "135/39476", "elapsed_time": "2h 13m 55s", "remaining_time": "27d 2h 23m 21s", "memory(GiB)": 38.53, "train_speed(s/it)": 59.515535} +{"loss": 0.37397903, "grad_norm": 0.27949998, "learning_rate": 7.09e-06, "load_balancing_loss": 1.14273572, "iteration": "140/39476", "elapsed_time": "2h 18m 28s", "remaining_time": "27d 0h 26m 10s", "memory(GiB)": 38.53, "train_speed(s/it)": 59.344378} +{"loss": 0.45193875, "grad_norm": 0.40383759, "learning_rate": 7.35e-06, "load_balancing_loss": 1.13927102, "iteration": "145/39476", "elapsed_time": "2h 22m 37s", "remaining_time": "26d 20h 42m 41s", "memory(GiB)": 38.53, "train_speed(s/it)": 59.010993} +{"loss": 0.38275439, "grad_norm": 0.26023218, "learning_rate": 7.6e-06, "load_balancing_loss": 1.16037714, "iteration": "150/39476", "elapsed_time": "2h 26m 57s", "remaining_time": "26d 18h 6m 31s", "memory(GiB)": 38.53, "train_speed(s/it)": 58.780232} +{"loss": 0.41769686, "grad_norm": 0.36750576, "learning_rate": 7.85e-06, "load_balancing_loss": 1.15820587, "iteration": "155/39476", "elapsed_time": "2h 31m 58s", "remaining_time": "26d 18h 32m 6s", "memory(GiB)": 38.53, "train_speed(s/it)": 58.826732} +{"loss": 0.37395322, "grad_norm": 0.4001134, "learning_rate": 8.11e-06, "load_balancing_loss": 1.15244615, "iteration": "160/39476", "elapsed_time": "2h 36m 35s", "remaining_time": "26d 17h 15m 7s", "memory(GiB)": 38.53, "train_speed(s/it)": 58.716722} +{"loss": 0.36512136, "grad_norm": 0.31565142, "learning_rate": 8.36e-06, "load_balancing_loss": 1.14771867, "iteration": "165/39476", "elapsed_time": "2h 41m 41s", "remaining_time": "26d 18h 1m 40s", "memory(GiB)": 38.53, "train_speed(s/it)": 58.795246} +{"loss": 0.37348658, "grad_norm": 0.24766535, "learning_rate": 8.61e-06, "load_balancing_loss": 1.1622498, "iteration": "170/39476", "elapsed_time": "2h 46m 33s", "remaining_time": "26d 17h 49m 46s", "memory(GiB)": 38.64, "train_speed(s/it)": 58.784557} +{"loss": 0.33336264, "grad_norm": 0.46295482, "learning_rate": 8.87e-06, "load_balancing_loss": 1.14322293, "iteration": "175/39476", "elapsed_time": "2h 51m 28s", "remaining_time": "26d 17h 49m 6s", "memory(GiB)": 38.64, "train_speed(s/it)": 58.791014} +{"loss": 0.39086345, "grad_norm": 0.52660465, "learning_rate": 9.12e-06, "load_balancing_loss": 1.14651704, "iteration": "180/39476", "elapsed_time": "2h 56m 22s", "remaining_time": "26d 17h 42m 58s", "memory(GiB)": 38.64, "train_speed(s/it)": 58.789132} +{"loss": 0.37840813, "grad_norm": 0.41599384, "learning_rate": 9.37e-06, "load_balancing_loss": 1.14309847, "iteration": "185/39476", "elapsed_time": "3h 1m 39s", "remaining_time": "26d 18h 58m 45s", "memory(GiB)": 38.64, "train_speed(s/it)": 58.912354} diff --git a/posttrain_models/sft_qwen3vl_4b_46k_multi_32k_timelens_1epoch_w_cot/v4-20260402-070604/args.json b/posttrain_models/sft_qwen3vl_4b_46k_multi_32k_timelens_1epoch_w_cot/v4-20260402-070604/args.json new file mode 100644 index 0000000000000000000000000000000000000000..b21c501cb875f2c3e15d6ff00c961a84120f9672 --- /dev/null +++ b/posttrain_models/sft_qwen3vl_4b_46k_multi_32k_timelens_1epoch_w_cot/v4-20260402-070604/args.json @@ -0,0 +1,389 @@ +{ + "output_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3vl_4b_46k_multi_32k_timelens_1epoch_w_cot/v4-20260402-070604", + "overwrite_output_dir": false, + "do_train": false, + "do_eval": false, + "do_predict": false, + "eval_strategy": "no", + "prediction_loss_only": false, + "per_device_train_batch_size": 4, + "per_device_eval_batch_size": 4, + "per_gpu_train_batch_size": null, + "per_gpu_eval_batch_size": null, + "gradient_accumulation_steps": 4, + "eval_accumulation_steps": null, + "eval_delay": 0, + "torch_empty_cache_steps": null, + "learning_rate": 1e-05, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "max_grad_norm": 1.0, + "num_train_epochs": 1.0, + "max_steps": -1, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_ratio": 0.03, + "warmup_steps": 0, + "log_level": "passive", + "log_level_replica": "warning", + "log_on_each_node": true, + "logging_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3vl_4b_46k_multi_32k_timelens_1epoch_w_cot/v4-20260402-070604/runs", + "logging_strategy": "steps", + "logging_first_step": true, + "logging_steps": 5, + "logging_nan_inf_filter": true, + "save_strategy": "epoch", + "save_steps": 500, + "save_total_limit": 1, + "save_safetensors": true, + "save_on_each_node": false, + "save_only_model": false, + "restore_callback_states_from_checkpoint": false, + "no_cuda": false, + "use_cpu": false, + "use_mps_device": false, + "seed": 42, + "data_seed": 42, + "jit_mode_eval": false, + "bf16": true, + "fp16": false, + "fp16_opt_level": "O1", + "half_precision_backend": "auto", + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": null, + "local_rank": 0, + "ddp_backend": null, + "tpu_num_cores": null, + "tpu_metrics_debug": false, + "debug": null, + "dataloader_drop_last": false, + "eval_steps": null, + "dataloader_num_workers": 24, + "dataloader_prefetch_factor": 2, + "past_index": -1, + "run_name": "sft_qwen3vl_4b_46k_multi_32k_timelens_1epoch_w_cot", + "disable_tqdm": null, + "remove_unused_columns": true, + "label_names": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "fsdp": [], + "fsdp_min_num_params": 0, + "fsdp_config": null, + "fsdp_transformer_layer_cls_to_wrap": null, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "deepspeed": { + "fp16": { + "enabled": "auto", + "loss_scale": 0, + "loss_scale_window": 1000, + "initial_scale_power": 16, + "hysteresis": 2, + "min_loss_scale": 1 + }, + "bf16": { + "enabled": "auto" + }, + "zero_optimization": { + "stage": 2, + "offload_optimizer": { + "device": "none", + "pin_memory": true + }, + "allgather_partitions": true, + "allgather_bucket_size": 200000000.0, + "overlap_comm": false, + "reduce_scatter": true, + "reduce_bucket_size": 200000000.0, + "contiguous_gradients": true + }, + "gradient_accumulation_steps": "auto", + "gradient_clipping": "auto", + "steps_per_print": 2000, + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "wall_clock_breakdown": false + }, + "label_smoothing_factor": 0.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "adafactor": false, + "group_by_length": false, + "length_column_name": "length", + "report_to": [ + "wandb" + ], + "project": "huggingface", + "trackio_space_id": "trackio", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": false, + "skip_memory_metrics": true, + "use_legacy_prediction_loop": false, + "push_to_hub": false, + "resume_from_checkpoint": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_token": null, + "hub_private_repo": null, + "hub_always_push": false, + "hub_revision": null, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "include_inputs_for_metrics": false, + "include_for_metrics": [], + "eval_do_concat_batches": true, + "fp16_backend": "auto", + "push_to_hub_model_id": null, + "push_to_hub_organization": null, + "push_to_hub_token": null, + "mp_parameters": "", + "auto_find_batch_size": false, + "full_determinism": false, + "torchdynamo": null, + "ray_scope": "last", + "ddp_timeout": 18000000, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "include_tokens_per_second": false, + "include_num_input_tokens_seen": false, + "neftune_noise_alpha": null, + "optim_target_modules": null, + "batch_eval_metrics": false, + "eval_on_start": false, + "use_liger_kernel": false, + "liger_kernel_config": null, + "eval_use_gather_object": false, + "average_tokens_across_devices": true, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": null, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "metric": [ + "temporal_grounding" + ], + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "Qwen/Qwen3-VL-4B-Instruct", + "model_type": "qwen3_vl", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": "flash_attn", + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_vl", + "system": null, + "max_length": 262144, + "truncation_strategy": "delete", + "max_pixels": null, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_free": false, + "padding_side": "right", + "loss_scale": "default", + "sequence_parallel_size": 4, + "response_prefix": null, + "template_backend": "swift", + "dataset": [ + "data/cosmo-cap/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "data/moment10m/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "data/vtimellm/train_filtered/train_pipeline_multi_refined_sft_with_cot.jsonl", + "data/timelens-100k/train/train_w_cot_reasoning_32k.jsonl" + ], + "val_dataset": [], + "cached_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 1, + "load_from_cache_file": true, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "ckpt_dir": null, + "lora_modules": [], + "train_type": "full", + "adapters": [], + "external_plugins": [], + "model_kwargs": {}, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "custom_register_path": [], + "use_hf": true, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [ + "model.visual.merger", + "model.visual.deepstack_merger_list" + ], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": false, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": null, + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_lark_webhook_url": null, + "swanlab_lark_secret": null, + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "early_stop_interval": null, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3-VL-4B-Instruct", + "model_info": "ModelInfo(model_type='qwen3_vl', model_dir='/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling={'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default'}, is_moe_model=False, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_vl', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct', hf_model_id='Qwen/Qwen3-VL-2B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking', hf_model_id='Qwen/Qwen3-VL-2B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct', hf_model_id='Qwen/Qwen3-VL-4B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking', hf_model_id='Qwen/Qwen3-VL-4B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct', hf_model_id='Qwen/Qwen3-VL-8B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking', hf_model_id='Qwen/Qwen3-VL-8B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct', hf_model_id='Qwen/Qwen3-VL-32B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking', hf_model_id='Qwen/Qwen3-VL-32B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='qwen3_vl', get_function=, model_arch=MultiModelKeys(arch_name='qwen3_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model'], aligner=['model.visual.merger', 'model.visual.deepstack_merger_list'], vision_tower=['model.visual'], generator=[]), architectures=['Qwen3VLForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, is_reranker=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])", + "model_dir": "/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17", + "_val_dataset_exists": [], + "hub": "", + "evaluation_strategy": "epoch", + "training_args": "Seq2SeqTrainingArguments(output_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3vl_4b_46k_multi_32k_timelens_1epoch_w_cot/v4-20260402-070604', overwrite_output_dir=False, do_train=False, do_eval=False, do_predict=False, eval_strategy=, prediction_loss_only=False, per_device_train_batch_size=4, per_device_eval_batch_size=4, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=4, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=1e-05, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=1.0, max_steps=-1, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_ratio=0.03, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3vl_4b_46k_multi_32k_timelens_1epoch_w_cot/v4-20260402-070604/runs', logging_strategy=, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=, save_steps=500, save_total_limit=1, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, jit_mode_eval=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=None, dataloader_num_workers=24, dataloader_prefetch_factor=2, past_index=-1, run_name='sft_qwen3vl_4b_46k_multi_32k_timelens_1epoch_w_cot', disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, deepspeed={'fp16': {'enabled': 'auto', 'loss_scale': 0, 'loss_scale_window': 1000, 'initial_scale_power': 16, 'hysteresis': 2, 'min_loss_scale': 1}, 'bf16': {'enabled': 'auto'}, 'zero_optimization': {'stage': 2, 'offload_optimizer': {'device': 'none', 'pin_memory': True}, 'allgather_partitions': True, 'allgather_bucket_size': 200000000.0, 'overlap_comm': False, 'reduce_scatter': True, 'reduce_bucket_size': 200000000.0, 'contiguous_gradients': True}, 'gradient_accumulation_steps': 'auto', 'gradient_clipping': 'auto', 'steps_per_print': 2000, 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'wall_clock_breakdown': False}, label_smoothing_factor=0.0, optim=, optim_args=None, adafactor=False, group_by_length=False, length_column_name='length', report_to=['wandb'], project='huggingface', trackio_space_id='trackio', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint=None, hub_model_id=None, hub_strategy=, hub_token=None, hub_private_repo=None, hub_always_push=False, hub_revision=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=18000000, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, include_tokens_per_second=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, liger_kernel_config=None, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=True, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, check_model=True, acc_strategy='token', train_dataloader_shuffle=True, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, metric=['temporal_grounding'], eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, use_flash_ckpt=False, sft_alpha=0, chord_sft_dataset=[], chord_sft_per_device_train_batch_size=None, chord_enable_phi_function=False, chord_mu_warmup_steps=None, chord_mu_decay_steps=None, chord_mu_peak=None, chord_mu_valley=None, train_type='full', local_repo_path=None, galore_config=None, padding_side='right', padding_free=False, task_type='causal_lm', problem_type=None)" +} \ No newline at end of file diff --git a/posttrain_models/sft_qwen3vl_4b_46k_multi_32k_timelens_1epoch_w_cot/v4-20260402-070604/logging.jsonl b/posttrain_models/sft_qwen3vl_4b_46k_multi_32k_timelens_1epoch_w_cot/v4-20260402-070604/logging.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..6de7f9913328bb68ef91ff1dd171d15ba192bfcb --- /dev/null +++ b/posttrain_models/sft_qwen3vl_4b_46k_multi_32k_timelens_1epoch_w_cot/v4-20260402-070604/logging.jsonl @@ -0,0 +1 @@ +{"model_parameter_info": "Qwen3VLForConditionalGeneration: 4437.8158M Params (4131.5732M Trainable [93.0993%]), 0.0001M Buffers.", "last_model_checkpoint": null, "best_model_checkpoint": null, "best_metric": null, "global_step": 0, "log_history": [], "memory": null} diff --git a/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/args.json b/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/args.json new file mode 100644 index 0000000000000000000000000000000000000000..8cd8dbfc19136fd397c69bf8df202a6f7a4a235a --- /dev/null +++ b/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/args.json @@ -0,0 +1,401 @@ +{ + "output_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021", + "overwrite_output_dir": false, + "do_train": false, + "do_eval": false, + "do_predict": false, + "eval_strategy": "no", + "prediction_loss_only": false, + "per_device_train_batch_size": 4, + "per_device_eval_batch_size": 4, + "per_gpu_train_batch_size": null, + "per_gpu_eval_batch_size": null, + "gradient_accumulation_steps": 4, + "eval_accumulation_steps": null, + "eval_delay": 0, + "torch_empty_cache_steps": null, + "learning_rate": 1e-05, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "max_grad_norm": 1.0, + "num_train_epochs": 1.0, + "max_steps": -1, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_ratio": 0.03, + "warmup_steps": 0, + "log_level": "passive", + "log_level_replica": "warning", + "log_on_each_node": true, + "logging_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/runs", + "logging_strategy": "steps", + "logging_first_step": true, + "logging_steps": 5, + "logging_nan_inf_filter": true, + "save_strategy": "epoch", + "save_steps": 500, + "save_total_limit": 1, + "save_safetensors": true, + "save_on_each_node": false, + "save_only_model": false, + "restore_callback_states_from_checkpoint": false, + "no_cuda": false, + "use_cpu": false, + "use_mps_device": false, + "seed": 42, + "data_seed": 42, + "jit_mode_eval": false, + "bf16": true, + "fp16": false, + "fp16_opt_level": "O1", + "half_precision_backend": "auto", + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": null, + "local_rank": 0, + "ddp_backend": null, + "tpu_num_cores": null, + "tpu_metrics_debug": false, + "debug": null, + "dataloader_drop_last": false, + "eval_steps": null, + "dataloader_num_workers": 24, + "dataloader_prefetch_factor": 2, + "past_index": -1, + "run_name": "sft_qwen3vl_4b_ntr_original_data_1epoch", + "disable_tqdm": null, + "remove_unused_columns": true, + "label_names": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "fsdp": [], + "fsdp_min_num_params": 0, + "fsdp_config": null, + "fsdp_transformer_layer_cls_to_wrap": null, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "deepspeed": { + "fp16": { + "enabled": "auto", + "loss_scale": 0, + "loss_scale_window": 1000, + "initial_scale_power": 16, + "hysteresis": 2, + "min_loss_scale": 1 + }, + "bf16": { + "enabled": "auto" + }, + "zero_optimization": { + "stage": 2, + "offload_optimizer": { + "device": "none", + "pin_memory": true + }, + "allgather_partitions": true, + "allgather_bucket_size": 200000000.0, + "overlap_comm": false, + "reduce_scatter": true, + "reduce_bucket_size": 200000000.0, + "contiguous_gradients": true + }, + "gradient_accumulation_steps": "auto", + "gradient_clipping": "auto", + "steps_per_print": 2000, + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "wall_clock_breakdown": false + }, + "label_smoothing_factor": 0.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "adafactor": false, + "group_by_length": false, + "length_column_name": "length", + "report_to": [ + "tensorboard" + ], + "project": "huggingface", + "trackio_space_id": "trackio", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": false, + "skip_memory_metrics": true, + "use_legacy_prediction_loop": false, + "push_to_hub": false, + "resume_from_checkpoint": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_token": null, + "hub_private_repo": null, + "hub_always_push": false, + "hub_revision": null, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "include_inputs_for_metrics": false, + "include_for_metrics": [], + "eval_do_concat_batches": true, + "fp16_backend": "auto", + "push_to_hub_model_id": null, + "push_to_hub_organization": null, + "push_to_hub_token": null, + "mp_parameters": "", + "auto_find_batch_size": false, + "full_determinism": false, + "torchdynamo": null, + "ray_scope": "last", + "ddp_timeout": 18000000, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "include_tokens_per_second": false, + "include_num_input_tokens_seen": false, + "neftune_noise_alpha": null, + "optim_target_modules": null, + "batch_eval_metrics": false, + "eval_on_start": false, + "use_liger_kernel": false, + "liger_kernel_config": null, + "eval_use_gather_object": false, + "average_tokens_across_devices": true, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": false, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "safe_serialization": true, + "max_shard_size": "5GB", + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "eval_metric": null, + "callbacks": [], + "early_stop_interval": null, + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "tuner_type": "full", + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "Qwen/Qwen3-VL-4B-Instruct", + "model_type": "qwen3_vl", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": null, + "experts_impl": null, + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_vl", + "system": null, + "max_length": 262144, + "truncation_strategy": "delete", + "max_pixels": null, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_side": "right", + "padding_free": false, + "loss_scale": "default", + "sequence_parallel_size": 4, + "template_backend": "swift", + "response_prefix": null, + "enable_thinking": null, + "add_non_thinking_prefix": true, + "dataset": [ + "/mnt/bn/strategy-mllm-train/user/tanyue/datasets/temp_data/original_data/train.jsonl" + ], + "val_dataset": [], + "cached_dataset": [], + "cached_val_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 1, + "load_from_cache_file": true, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "disable_auto_column_mapping": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "structured_outputs_regex": null, + "adapters": [], + "external_plugins": [], + "custom_register_path": [], + "model_kwargs": {}, + "enable_npu_model_patch": true, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "use_hf": true, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [ + "model.visual.merger", + "model.visual.deepstack_merger_list" + ], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": false, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": "ms-swift", + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_notification_method": null, + "swanlab_webhook_url": null, + "swanlab_secret": null, + "swanlab_sender_email": null, + "swanlab_receiver_email": null, + "swanlab_smtp_server": null, + "swanlab_smtp_port": null, + "swanlab_email_language": "zh", + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "swift_version": "4.2.1", + "ckpt_dir": null, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3-VL-4B-Instruct", + "model_info": "ModelInfo(model_type='qwen3_vl', model_dir='/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling={'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default'}, is_moe_model=False, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_vl', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct', hf_model_id='Qwen/Qwen3-VL-2B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking', hf_model_id='Qwen/Qwen3-VL-2B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct', hf_model_id='Qwen/Qwen3-VL-4B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking', hf_model_id='Qwen/Qwen3-VL-4B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct', hf_model_id='Qwen/Qwen3-VL-8B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking', hf_model_id='Qwen/Qwen3-VL-8B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct', hf_model_id='Qwen/Qwen3-VL-32B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking', hf_model_id='Qwen/Qwen3-VL-32B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_vl', ignore_patterns=None, requires=None, tags=[])], loader=, template='qwen3_vl', model_arch=MultiModelKeys(arch_name='qwen3_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model', 'lm_head'], aligner=['model.visual.merger', 'model.visual.deepstack_merger_list'], vision_tower=['model.visual'], generator=[]), mcore_model_type=None, architectures=['Qwen3VLForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57', 'qwen_vl_utils>=0.0.14', 'decord'], tags=[])", + "model_dir": "/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17", + "template_meta": "QwenTemplateMeta(template_type='qwen3_vl', prefix=[], prompt=['<|im_start|>user\\n{{QUERY}}<|im_end|>\\n<|im_start|>assistant\\n'], chat_sep=['<|im_end|>\\n'], suffix=['<|im_end|>\\n'], template_cls=, system_prefix=['<|im_start|>system\\n{{SYSTEM}}<|im_end|>\\n'], default_system=None, auto_add_bos=False, stop_words=['<|endoftext|>'], agent_template='hermes', is_thinking=False, thinking_prefix='\\n', non_thinking_prefix='', history_thinking_prefix='')", + "_val_dataset_exists": false, + "hub": "", + "evaluation_strategy": "epoch", + "training_args": "Seq2SeqTrainingArguments(output_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021', overwrite_output_dir=False, do_train=False, do_eval=False, do_predict=False, eval_strategy=, prediction_loss_only=False, per_device_train_batch_size=4, per_device_eval_batch_size=4, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=4, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=1e-05, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=1.0, max_steps=-1, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_ratio=0.03, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/runs', logging_strategy=, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=, save_steps=500, save_total_limit=1, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, jit_mode_eval=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=None, dataloader_num_workers=24, dataloader_prefetch_factor=2, past_index=-1, run_name='sft_qwen3vl_4b_ntr_original_data_1epoch', disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, deepspeed={'fp16': {'enabled': 'auto', 'loss_scale': 0, 'loss_scale_window': 1000, 'initial_scale_power': 16, 'hysteresis': 2, 'min_loss_scale': 1}, 'bf16': {'enabled': 'auto'}, 'zero_optimization': {'stage': 2, 'offload_optimizer': {'device': 'none', 'pin_memory': True}, 'allgather_partitions': True, 'allgather_bucket_size': 200000000.0, 'overlap_comm': False, 'reduce_scatter': True, 'reduce_bucket_size': 200000000.0, 'contiguous_gradients': True}, 'gradient_accumulation_steps': 'auto', 'gradient_clipping': 'auto', 'steps_per_print': 2000, 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'wall_clock_breakdown': False}, label_smoothing_factor=0.0, optim=, optim_args=None, adafactor=False, group_by_length=False, length_column_name='length', report_to=['tensorboard'], project='huggingface', trackio_space_id='trackio', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint=None, hub_model_id=None, hub_strategy=, hub_token=None, hub_private_repo=None, hub_always_push=False, hub_revision=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=18000000, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, include_tokens_per_second=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, liger_kernel_config=None, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=False, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, safe_serialization=True, max_shard_size='5GB', check_model=True, acc_strategy='token', train_dataloader_shuffle=True, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, eval_metric=None, callbacks=[], early_stop_interval=None, eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, tuner_type='full', use_galore=False, galore_target_modules=None, galore_rank=128, galore_update_proj_gap=50, galore_scale=1.0, galore_proj_type='std', galore_optim_per_parameter=False, galore_with_embedding=False, galore_quantization=False, galore_proj_quant=False, galore_proj_bits=4, galore_proj_group_size=256, galore_cos_threshold=0.4, galore_gamma_proj=2, galore_queue_size=5, lisa_activated_layers=0, lisa_step_interval=20, use_flash_ckpt=False)" +} \ No newline at end of file diff --git a/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/checkpoint-418/added_tokens.json b/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/checkpoint-418/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..b54f9135e44c1e81047e8d05cb027af8bc039eed --- /dev/null +++ b/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/checkpoint-418/added_tokens.json @@ -0,0 +1,28 @@ +{ + "": 151668, + "": 151658, + "": 151666, + "": 151667, + "": 151657, + "": 151665, + "<|box_end|>": 151649, + "<|box_start|>": 151648, + "<|endoftext|>": 151643, + "<|file_sep|>": 151664, + "<|fim_middle|>": 151660, + "<|fim_pad|>": 151662, + "<|fim_prefix|>": 151659, + "<|fim_suffix|>": 151661, + "<|im_end|>": 151645, + "<|im_start|>": 151644, + "<|image_pad|>": 151655, + "<|object_ref_end|>": 151647, + "<|object_ref_start|>": 151646, + "<|quad_end|>": 151651, + "<|quad_start|>": 151650, + "<|repo_name|>": 151663, + "<|video_pad|>": 151656, + "<|vision_end|>": 151653, + "<|vision_pad|>": 151654, + "<|vision_start|>": 151652 +} diff --git a/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/checkpoint-418/args.json b/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/checkpoint-418/args.json new file mode 100644 index 0000000000000000000000000000000000000000..8cd8dbfc19136fd397c69bf8df202a6f7a4a235a --- /dev/null +++ b/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/checkpoint-418/args.json @@ -0,0 +1,401 @@ +{ + "output_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021", + "overwrite_output_dir": false, + "do_train": false, + "do_eval": false, + "do_predict": false, + "eval_strategy": "no", + "prediction_loss_only": false, + "per_device_train_batch_size": 4, + "per_device_eval_batch_size": 4, + "per_gpu_train_batch_size": null, + "per_gpu_eval_batch_size": null, + "gradient_accumulation_steps": 4, + "eval_accumulation_steps": null, + "eval_delay": 0, + "torch_empty_cache_steps": null, + "learning_rate": 1e-05, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "max_grad_norm": 1.0, + "num_train_epochs": 1.0, + "max_steps": -1, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_ratio": 0.03, + "warmup_steps": 0, + "log_level": "passive", + "log_level_replica": "warning", + "log_on_each_node": true, + "logging_dir": "/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/runs", + "logging_strategy": "steps", + "logging_first_step": true, + "logging_steps": 5, + "logging_nan_inf_filter": true, + "save_strategy": "epoch", + "save_steps": 500, + "save_total_limit": 1, + "save_safetensors": true, + "save_on_each_node": false, + "save_only_model": false, + "restore_callback_states_from_checkpoint": false, + "no_cuda": false, + "use_cpu": false, + "use_mps_device": false, + "seed": 42, + "data_seed": 42, + "jit_mode_eval": false, + "bf16": true, + "fp16": false, + "fp16_opt_level": "O1", + "half_precision_backend": "auto", + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": null, + "local_rank": 0, + "ddp_backend": null, + "tpu_num_cores": null, + "tpu_metrics_debug": false, + "debug": null, + "dataloader_drop_last": false, + "eval_steps": null, + "dataloader_num_workers": 24, + "dataloader_prefetch_factor": 2, + "past_index": -1, + "run_name": "sft_qwen3vl_4b_ntr_original_data_1epoch", + "disable_tqdm": null, + "remove_unused_columns": true, + "label_names": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "fsdp": [], + "fsdp_min_num_params": 0, + "fsdp_config": null, + "fsdp_transformer_layer_cls_to_wrap": null, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "deepspeed": { + "fp16": { + "enabled": "auto", + "loss_scale": 0, + "loss_scale_window": 1000, + "initial_scale_power": 16, + "hysteresis": 2, + "min_loss_scale": 1 + }, + "bf16": { + "enabled": "auto" + }, + "zero_optimization": { + "stage": 2, + "offload_optimizer": { + "device": "none", + "pin_memory": true + }, + "allgather_partitions": true, + "allgather_bucket_size": 200000000.0, + "overlap_comm": false, + "reduce_scatter": true, + "reduce_bucket_size": 200000000.0, + "contiguous_gradients": true + }, + "gradient_accumulation_steps": "auto", + "gradient_clipping": "auto", + "steps_per_print": 2000, + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "wall_clock_breakdown": false + }, + "label_smoothing_factor": 0.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "adafactor": false, + "group_by_length": false, + "length_column_name": "length", + "report_to": [ + "tensorboard" + ], + "project": "huggingface", + "trackio_space_id": "trackio", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": false, + "skip_memory_metrics": true, + "use_legacy_prediction_loop": false, + "push_to_hub": false, + "resume_from_checkpoint": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_token": null, + "hub_private_repo": null, + "hub_always_push": false, + "hub_revision": null, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "include_inputs_for_metrics": false, + "include_for_metrics": [], + "eval_do_concat_batches": true, + "fp16_backend": "auto", + "push_to_hub_model_id": null, + "push_to_hub_organization": null, + "push_to_hub_token": null, + "mp_parameters": "", + "auto_find_batch_size": false, + "full_determinism": false, + "torchdynamo": null, + "ray_scope": "last", + "ddp_timeout": 18000000, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "include_tokens_per_second": false, + "include_num_input_tokens_seen": false, + "neftune_noise_alpha": null, + "optim_target_modules": null, + "batch_eval_metrics": false, + "eval_on_start": false, + "use_liger_kernel": false, + "liger_kernel_config": null, + "eval_use_gather_object": false, + "average_tokens_across_devices": true, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": false, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "safe_serialization": true, + "max_shard_size": "5GB", + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "eval_metric": null, + "callbacks": [], + "early_stop_interval": null, + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "tuner_type": "full", + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "Qwen/Qwen3-VL-4B-Instruct", + "model_type": "qwen3_vl", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": null, + "experts_impl": null, + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_vl", + "system": null, + "max_length": 262144, + "truncation_strategy": "delete", + "max_pixels": null, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_side": "right", + "padding_free": false, + "loss_scale": "default", + "sequence_parallel_size": 4, + "template_backend": "swift", + "response_prefix": null, + "enable_thinking": null, + "add_non_thinking_prefix": true, + "dataset": [ + "/mnt/bn/strategy-mllm-train/user/tanyue/datasets/temp_data/original_data/train.jsonl" + ], + "val_dataset": [], + "cached_dataset": [], + "cached_val_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 1, + "load_from_cache_file": true, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "disable_auto_column_mapping": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "structured_outputs_regex": null, + "adapters": [], + "external_plugins": [], + "custom_register_path": [], + "model_kwargs": {}, + "enable_npu_model_patch": true, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "use_hf": true, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [ + "model.visual.merger", + "model.visual.deepstack_merger_list" + ], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": false, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": "ms-swift", + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_notification_method": null, + "swanlab_webhook_url": null, + "swanlab_secret": null, + "swanlab_sender_email": null, + "swanlab_receiver_email": null, + "swanlab_smtp_server": null, + "swanlab_smtp_port": null, + "swanlab_email_language": "zh", + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "swift_version": "4.2.1", + "ckpt_dir": null, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3-VL-4B-Instruct", + "model_info": "ModelInfo(model_type='qwen3_vl', model_dir='/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling={'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default'}, is_moe_model=False, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_vl', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct', hf_model_id='Qwen/Qwen3-VL-2B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking', hf_model_id='Qwen/Qwen3-VL-2B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-2B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct', hf_model_id='Qwen/Qwen3-VL-4B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking', hf_model_id='Qwen/Qwen3-VL-4B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-4B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct', hf_model_id='Qwen/Qwen3-VL-8B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking', hf_model_id='Qwen/Qwen3-VL-8B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-8B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct', hf_model_id='Qwen/Qwen3-VL-32B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking', hf_model_id='Qwen/Qwen3-VL-32B-Thinking', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', hf_model_id='Qwen/Qwen3-VL-32B-Thinking-FP8', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_vl', ignore_patterns=None, requires=None, tags=[])], loader=, template='qwen3_vl', model_arch=MultiModelKeys(arch_name='qwen3_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model', 'lm_head'], aligner=['model.visual.merger', 'model.visual.deepstack_merger_list'], vision_tower=['model.visual'], generator=[]), mcore_model_type=None, architectures=['Qwen3VLForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, task_type=None, ignore_patterns=None, requires=['transformers>=4.57', 'qwen_vl_utils>=0.0.14', 'decord'], tags=[])", + "model_dir": "/home/tiger/.cache/huggingface/hub/models--Qwen--Qwen3-VL-4B-Instruct/snapshots/ebb281ec70b05090aa6165b016eac8ec08e71b17", + "template_meta": "QwenTemplateMeta(template_type='qwen3_vl', prefix=[], prompt=['<|im_start|>user\\n{{QUERY}}<|im_end|>\\n<|im_start|>assistant\\n'], chat_sep=['<|im_end|>\\n'], suffix=['<|im_end|>\\n'], template_cls=, system_prefix=['<|im_start|>system\\n{{SYSTEM}}<|im_end|>\\n'], default_system=None, auto_add_bos=False, stop_words=['<|endoftext|>'], agent_template='hermes', is_thinking=False, thinking_prefix='\\n', non_thinking_prefix='', history_thinking_prefix='')", + "_val_dataset_exists": false, + "hub": "", + "evaluation_strategy": "epoch", + "training_args": "Seq2SeqTrainingArguments(output_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021', overwrite_output_dir=False, do_train=False, do_eval=False, do_predict=False, eval_strategy=, prediction_loss_only=False, per_device_train_batch_size=4, per_device_eval_batch_size=4, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=4, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=1e-05, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=1.0, max_steps=-1, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_ratio=0.03, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='/mnt/bn/strategy-mllm-train/user/tanyue/models/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/runs', logging_strategy=, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=, save_steps=500, save_total_limit=1, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, jit_mode_eval=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=None, dataloader_num_workers=24, dataloader_prefetch_factor=2, past_index=-1, run_name='sft_qwen3vl_4b_ntr_original_data_1epoch', disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, deepspeed={'fp16': {'enabled': 'auto', 'loss_scale': 0, 'loss_scale_window': 1000, 'initial_scale_power': 16, 'hysteresis': 2, 'min_loss_scale': 1}, 'bf16': {'enabled': 'auto'}, 'zero_optimization': {'stage': 2, 'offload_optimizer': {'device': 'none', 'pin_memory': True}, 'allgather_partitions': True, 'allgather_bucket_size': 200000000.0, 'overlap_comm': False, 'reduce_scatter': True, 'reduce_bucket_size': 200000000.0, 'contiguous_gradients': True}, 'gradient_accumulation_steps': 'auto', 'gradient_clipping': 'auto', 'steps_per_print': 2000, 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'wall_clock_breakdown': False}, label_smoothing_factor=0.0, optim=, optim_args=None, adafactor=False, group_by_length=False, length_column_name='length', report_to=['tensorboard'], project='huggingface', trackio_space_id='trackio', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint=None, hub_model_id=None, hub_strategy=, hub_token=None, hub_private_repo=None, hub_always_push=False, hub_revision=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=18000000, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, include_tokens_per_second=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, liger_kernel_config=None, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=False, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, safe_serialization=True, max_shard_size='5GB', check_model=True, acc_strategy='token', train_dataloader_shuffle=True, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, eval_metric=None, callbacks=[], early_stop_interval=None, eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, tuner_type='full', use_galore=False, galore_target_modules=None, galore_rank=128, galore_update_proj_gap=50, galore_scale=1.0, galore_proj_type='std', galore_optim_per_parameter=False, galore_with_embedding=False, galore_quantization=False, galore_proj_quant=False, galore_proj_bits=4, galore_proj_group_size=256, galore_cos_threshold=0.4, galore_gamma_proj=2, galore_queue_size=5, lisa_activated_layers=0, lisa_step_interval=20, use_flash_ckpt=False)" +} \ No newline at end of file diff --git a/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/checkpoint-418/chat_template.jinja b/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/checkpoint-418/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..124386803f142761528f710e77ae483f5f8c4fc4 --- /dev/null +++ b/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/checkpoint-418/chat_template.jinja @@ -0,0 +1,120 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].role == 'system' %} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n\n' }} + {%- endif %} + {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }} +{%- else %} + {%- if messages[0].role == 'system' %} + {{- '<|im_start|>system\n' }} + {%- if messages[0].content is string %} + {{- messages[0].content }} + {%- else %} + {%- for content in messages[0].content %} + {%- if 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- for message in messages %} + {%- if message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role + '\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content_item in message.content %} + {%- if 'text' in content_item %} + {{- content_item.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {%- if message.tool_calls %} + {%- for tool_call in message.tool_calls %} + {%- if (loop.first and message.content) or (not loop.first) %} + {{- '\n' }} + {%- endif %} + {%- if tool_call.function %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {%- if tool_call.arguments is string %} + {{- tool_call.arguments }} + {%- else %} + {{- tool_call.arguments | tojson }} + {%- endif %} + {{- '}\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {%- if message.content is string %} + {{- message.content }} + {%- else %} + {%- for content in message.content %} + {%- if content.type == 'image' or 'image' in content or 'image_url' in content %} + {%- set image_count.value = image_count.value + 1 %} + {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%} + <|vision_start|><|image_pad|><|vision_end|> + {%- elif content.type == 'video' or 'video' in content %} + {%- set video_count.value = video_count.value + 1 %} + {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%} + <|vision_start|><|video_pad|><|vision_end|> + {%- elif 'text' in content %} + {{- content.text }} + {%- endif %} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} +{%- endif %} diff --git a/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/checkpoint-418/config.json b/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/checkpoint-418/config.json new file mode 100644 index 0000000000000000000000000000000000000000..b7fe733236ab4989a27e8f42fc4b3a53ea446988 --- /dev/null +++ b/posttrain_models/sft_qwen3vl_4b_ntr_original_data_1epoch/v0-20260518-163021/checkpoint-418/config.json @@ -0,0 +1,69 @@ +{ + "architectures": [ + "Qwen3VLForConditionalGeneration" + ], + "dtype": "bfloat16", + "eos_token_id": 151645, + "hidden_size": 2560, + "image_token_id": 151655, + "model_type": "qwen3_vl", + "pad_token_id": 151643, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 151643, + "dtype": "bfloat16", + "eos_token_id": 151645, + "head_dim": 128, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9728, + "max_position_embeddings": 262144, + "model_type": "qwen3_vl_text", + "num_attention_heads": 32, + "num_hidden_layers": 36, + "num_key_value_heads": 8, + "pad_token_id": 151643, + "rms_norm_eps": 1e-06, + "rope_scaling": { + "mrope_interleaved": true, + "mrope_section": [ + 24, + 20, + 20 + ], + "rope_type": "default" + }, + "rope_theta": 5000000, + "tie_word_embeddings": true, + "use_cache": false, + "vocab_size": 151936 + }, + "tie_word_embeddings": true, + "transformers_version": "4.57.0", + "video_token_id": 151656, + "vision_config": { + "deepstack_visual_indexes": [ + 5, + 11, + 17 + ], + "depth": 24, + "dtype": "bfloat16", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_vl", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652 +}