Upload continuous evaluation checkpoints, batch 26

.gitattributes

@@ -175,3 +175,6 @@ continuous/qwen3/siglip2_sm16_512/finetune/tokenizer.json filter=lfs diff=lfs me
 continuous/qwen3/webssl_dino1b_full2b_224/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 continuous/qwen3/webssl_mae1b_full2b_224/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 continuous/qwen3/webssl_mae300m_full2b_224/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+continuous/qwen3/webssl_mae3b_full2b_224/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+continuous/qwen3base/mc2_l14_224/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+continuous/qwen3base/siglip2_g16_384/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text

continuous/qwen3/webssl_mae3b_full2b_224/finetune/model-00002-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a88e597d90fa4bc1abdab054d18ae08cf4296514517a1e5cb3d63d3c17a31c0
+size 4441301232

continuous/qwen3/webssl_mae3b_full2b_224/finetune/model-00003-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f3a411e0277aa56ea24f1c5eada647e072a3273121db3d867d2571b74a45d34
+size 622329984

continuous/qwen3/webssl_mae3b_full2b_224/finetune/model.safetensors.index.json

@@ -0,0 +1,747 @@
+{
+  "metadata": {
+    "total_parameters": 4688552960,
+    "total_size": 10020415488
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00003-of-00003.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.11.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.11.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.12.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.12.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.13.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.13.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.14.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.14.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.15.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.15.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.16.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.16.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.17.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.17.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.18.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.18.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.19.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.19.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.20.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.20.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.21.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.21.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.22.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.22.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.23.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.23.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.24.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.24.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.25.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.25.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.26.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.26.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.27.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.27.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "model.mm_projector.0.bias": "model-00002-of-00003.safetensors",
+    "model.mm_projector.0.weight": "model-00002-of-00003.safetensors",
+    "model.mm_projector.2.bias": "model-00002-of-00003.safetensors",
+    "model.mm_projector.2.weight": "model-00002-of-00003.safetensors",
+    "model.norm.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.embeddings.cls_token": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.embeddings.patch_embeddings.projection.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.embeddings.patch_embeddings.projection.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.embeddings.position_embeddings": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.attention.attention.key.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.attention.attention.key.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.attention.attention.query.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.attention.attention.query.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.attention.attention.value.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.attention.attention.value.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.attention.output.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.attention.output.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.intermediate.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.intermediate.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.layernorm_after.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.layernorm_after.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.layernorm_before.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.layernorm_before.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.output.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.0.output.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.attention.attention.key.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.attention.attention.key.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.attention.attention.query.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.attention.attention.query.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.attention.attention.value.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.attention.attention.value.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.attention.output.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.attention.output.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.intermediate.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.intermediate.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.layernorm_after.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.layernorm_after.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.layernorm_before.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.layernorm_before.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.output.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.1.output.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.10.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.11.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.12.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.13.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.14.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.15.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.16.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.17.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.18.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.19.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.attention.attention.key.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.attention.attention.key.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.attention.attention.query.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.attention.attention.query.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.attention.attention.value.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.attention.attention.value.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.attention.output.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.attention.output.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.intermediate.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.intermediate.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.layernorm_after.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.layernorm_after.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.layernorm_before.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.layernorm_before.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.output.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.2.output.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.20.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.21.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.22.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.23.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.24.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.25.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.attention.attention.key.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.attention.attention.key.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.attention.attention.query.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.attention.attention.query.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.attention.attention.value.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.attention.attention.value.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.attention.output.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.attention.output.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.intermediate.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.intermediate.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.layernorm_after.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.layernorm_after.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.layernorm_before.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.layernorm_before.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.output.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.3.output.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.attention.attention.key.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.attention.attention.key.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.attention.attention.query.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.attention.attention.query.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.attention.attention.value.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.attention.attention.value.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.attention.output.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.attention.output.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.intermediate.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.intermediate.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.layernorm_after.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.layernorm_after.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.layernorm_before.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.layernorm_before.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.output.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.4.output.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.attention.attention.key.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.attention.attention.key.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.attention.attention.query.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.attention.attention.query.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.attention.attention.value.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.attention.attention.value.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.attention.output.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.attention.output.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.intermediate.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.intermediate.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.layernorm_after.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.layernorm_after.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.layernorm_before.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.layernorm_before.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.output.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.5.output.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.attention.attention.key.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.attention.attention.key.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.attention.attention.query.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.attention.attention.query.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.attention.attention.value.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.attention.attention.value.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.attention.output.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.attention.output.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.intermediate.dense.bias": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.intermediate.dense.weight": "model-00001-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.6.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.7.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.8.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.attention.attention.key.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.attention.attention.key.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.attention.attention.query.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.attention.attention.query.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.attention.attention.value.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.attention.attention.value.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.attention.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.attention.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.intermediate.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.intermediate.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.layernorm_after.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.layernorm_after.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.layernorm_before.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.layernorm_before.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.output.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.encoder.layer.9.output.dense.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.layernorm.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.layernorm.weight": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.pooler.dense.bias": "model-00002-of-00003.safetensors",
+    "model.vision_tower.vision_tower.pooler.dense.weight": "model-00002-of-00003.safetensors"
+  }
+}

continuous/qwen3/webssl_mae3b_full2b_224/finetune/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
+  }
+}

continuous/qwen3/webssl_mae3b_full2b_224/finetune/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
+size 11422654

continuous/qwen3/webssl_mae3b_full2b_224/finetune/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": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "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": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "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": 8192,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

continuous/qwen3/webssl_mae3b_full2b_224/finetune/trainer_state.json

@@ -0,0 +1,1856 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999037443449803,
+  "eval_steps": 500,
+  "global_step": 2597,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003850226200789296,
+      "grad_norm": 15.974503517150879,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 2.303,
+      "step": 10
+    },
+    {
+      "epoch": 0.007700452401578592,
+      "grad_norm": 6.52081298828125,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.9699,
+      "step": 20
+    },
+    {
+      "epoch": 0.01155067860236789,
+      "grad_norm": 4.452016830444336,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.5767,
+      "step": 30
+    },
+    {
+      "epoch": 0.015400904803157185,
+      "grad_norm": 1.2203242778778076,
+      "learning_rate": 5e-06,
+      "loss": 1.2186,
+      "step": 40
+    },
+    {
+      "epoch": 0.019251131003946483,
+      "grad_norm": 0.7950222492218018,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.0897,
+      "step": 50
+    },
+    {
+      "epoch": 0.02310135720473578,
+      "grad_norm": 0.43575164675712585,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.9989,
+      "step": 60
+    },
+    {
+      "epoch": 0.026951583405525073,
+      "grad_norm": 0.32255637645721436,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.9452,
+      "step": 70
+    },
+    {
+      "epoch": 0.03080180960631437,
+      "grad_norm": 0.3070615828037262,
+      "learning_rate": 9.999996111488723e-06,
+      "loss": 0.9434,
+      "step": 80
+    },
+    {
+      "epoch": 0.034652035807103666,
+      "grad_norm": 0.28436416387557983,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 0.9107,
+      "step": 90
+    },
+    {
+      "epoch": 0.038502262007892966,
+      "grad_norm": 0.2579933702945709,
+      "learning_rate": 9.998285264324193e-06,
+      "loss": 0.8931,
+      "step": 100
+    },
+    {
+      "epoch": 0.04235248820868226,
+      "grad_norm": 0.26444143056869507,
+      "learning_rate": 9.996263605626052e-06,
+      "loss": 0.8743,
+      "step": 110
+    },
+    {
+      "epoch": 0.04620271440947156,
+      "grad_norm": 0.26989179849624634,
+      "learning_rate": 9.99346483580503e-06,
+      "loss": 0.8852,
+      "step": 120
+    },
+    {
+      "epoch": 0.05005294061026085,
+      "grad_norm": 0.2534806430339813,
+      "learning_rate": 9.989889390177467e-06,
+      "loss": 0.8807,
+      "step": 130
+    },
+    {
+      "epoch": 0.053903166811050146,
+      "grad_norm": 0.29042115807533264,
+      "learning_rate": 9.985537824862647e-06,
+      "loss": 0.8749,
+      "step": 140
+    },
+    {
+      "epoch": 0.057753393011839446,
+      "grad_norm": 0.30993881821632385,
+      "learning_rate": 9.980410816696317e-06,
+      "loss": 0.8605,
+      "step": 150
+    },
+    {
+      "epoch": 0.06160361921262874,
+      "grad_norm": 0.28218474984169006,
+      "learning_rate": 9.97450916312541e-06,
+      "loss": 0.8656,
+      "step": 160
+    },
+    {
+      "epoch": 0.06545384541341803,
+      "grad_norm": 0.2574910819530487,
+      "learning_rate": 9.967833782084004e-06,
+      "loss": 0.8656,
+      "step": 170
+    },
+    {
+      "epoch": 0.06930407161420733,
+      "grad_norm": 0.27212774753570557,
+      "learning_rate": 9.960385711850554e-06,
+      "loss": 0.8933,
+      "step": 180
+    },
+    {
+      "epoch": 0.07315429781499663,
+      "grad_norm": 0.2596517503261566,
+      "learning_rate": 9.952166110886394e-06,
+      "loss": 0.8832,
+      "step": 190
+    },
+    {
+      "epoch": 0.07700452401578593,
+      "grad_norm": 0.308794766664505,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.8729,
+      "step": 200
+    },
+    {
+      "epoch": 0.08085475021657522,
+      "grad_norm": 0.26154428720474243,
+      "learning_rate": 9.933417550425952e-06,
+      "loss": 0.8667,
+      "step": 210
+    },
+    {
+      "epoch": 0.08470497641736452,
+      "grad_norm": 0.26733750104904175,
+      "learning_rate": 9.9228915070518e-06,
+      "loss": 0.8561,
+      "step": 220
+    },
+    {
+      "epoch": 0.08855520261815382,
+      "grad_norm": 0.2713722586631775,
+      "learning_rate": 9.911599764737632e-06,
+      "loss": 0.8559,
+      "step": 230
+    },
+    {
+      "epoch": 0.09240542881894312,
+      "grad_norm": 0.27083706855773926,
+      "learning_rate": 9.899544079783606e-06,
+      "loss": 0.8502,
+      "step": 240
+    },
+    {
+      "epoch": 0.0962556550197324,
+      "grad_norm": 0.30298346281051636,
+      "learning_rate": 9.886726327312338e-06,
+      "loss": 0.8677,
+      "step": 250
+    },
+    {
+      "epoch": 0.1001058812205217,
+      "grad_norm": 0.2539936602115631,
+      "learning_rate": 9.87314850097724e-06,
+      "loss": 0.8644,
+      "step": 260
+    },
+    {
+      "epoch": 0.103956107421311,
+      "grad_norm": 0.2589256465435028,
+      "learning_rate": 9.858812712652452e-06,
+      "loss": 0.8517,
+      "step": 270
+    },
+    {
+      "epoch": 0.10780633362210029,
+      "grad_norm": 0.25837773084640503,
+      "learning_rate": 9.843721192104341e-06,
+      "loss": 0.852,
+      "step": 280
+    },
+    {
+      "epoch": 0.11165655982288959,
+      "grad_norm": 0.27550560235977173,
+      "learning_rate": 9.827876286644698e-06,
+      "loss": 0.8692,
+      "step": 290
+    },
+    {
+      "epoch": 0.11550678602367889,
+      "grad_norm": 0.23211924731731415,
+      "learning_rate": 9.811280460765643e-06,
+      "loss": 0.8287,
+      "step": 300
+    },
+    {
+      "epoch": 0.11935701222446819,
+      "grad_norm": 0.2618008852005005,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.8433,
+      "step": 310
+    },
+    {
+      "epoch": 0.12320723842525748,
+      "grad_norm": 0.26063844561576843,
+      "learning_rate": 9.775846489301279e-06,
+      "loss": 0.8436,
+      "step": 320
+    },
+    {
+      "epoch": 0.12705746462604678,
+      "grad_norm": 0.25660184025764465,
+      "learning_rate": 9.757013855061157e-06,
+      "loss": 0.8432,
+      "step": 330
+    },
+    {
+      "epoch": 0.13090769082683607,
+      "grad_norm": 0.26929181814193726,
+      "learning_rate": 9.737441322234758e-06,
+      "loss": 0.8533,
+      "step": 340
+    },
+    {
+      "epoch": 0.13475791702762538,
+      "grad_norm": 0.2817429006099701,
+      "learning_rate": 9.717131935103605e-06,
+      "loss": 0.8534,
+      "step": 350
+    },
+    {
+      "epoch": 0.13860814322841467,
+      "grad_norm": 0.25834691524505615,
+      "learning_rate": 9.696088852558398e-06,
+      "loss": 0.8395,
+      "step": 360
+    },
+    {
+      "epoch": 0.14245836942920398,
+      "grad_norm": 0.2706683576107025,
+      "learning_rate": 9.674315347607686e-06,
+      "loss": 0.8664,
+      "step": 370
+    },
+    {
+      "epoch": 0.14630859562999327,
+      "grad_norm": 0.25769925117492676,
+      "learning_rate": 9.651814806868796e-06,
+      "loss": 0.827,
+      "step": 380
+    },
+    {
+      "epoch": 0.15015882183078255,
+      "grad_norm": 0.27218571305274963,
+      "learning_rate": 9.62859073004107e-06,
+      "loss": 0.8345,
+      "step": 390
+    },
+    {
+      "epoch": 0.15400904803157187,
+      "grad_norm": 0.246397465467453,
+      "learning_rate": 9.604646729361547e-06,
+      "loss": 0.8383,
+      "step": 400
+    },
+    {
+      "epoch": 0.15785927423236115,
+      "grad_norm": 0.2500905394554138,
+      "learning_rate": 9.579986529043099e-06,
+      "loss": 0.8263,
+      "step": 410
+    },
+    {
+      "epoch": 0.16170950043315044,
+      "grad_norm": 0.2492186725139618,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.8421,
+      "step": 420
+    },
+    {
+      "epoch": 0.16555972663393975,
+      "grad_norm": 0.2621612548828125,
+      "learning_rate": 9.528532982727281e-06,
+      "loss": 0.8315,
+      "step": 430
+    },
+    {
+      "epoch": 0.16940995283472904,
+      "grad_norm": 0.2424831986427307,
+      "learning_rate": 9.501747639735018e-06,
+      "loss": 0.8302,
+      "step": 440
+    },
+    {
+      "epoch": 0.17326017903551832,
+      "grad_norm": 0.28034359216690063,
+      "learning_rate": 9.474262101869272e-06,
+      "loss": 0.8195,
+      "step": 450
+    },
+    {
+      "epoch": 0.17711040523630764,
+      "grad_norm": 0.26365792751312256,
+      "learning_rate": 9.446080644188141e-06,
+      "loss": 0.8247,
+      "step": 460
+    },
+    {
+      "epoch": 0.18096063143709692,
+      "grad_norm": 0.263776957988739,
+      "learning_rate": 9.417207649992017e-06,
+      "loss": 0.8337,
+      "step": 470
+    },
+    {
+      "epoch": 0.18481085763788624,
+      "grad_norm": 0.2565253973007202,
+      "learning_rate": 9.387647610141816e-06,
+      "loss": 0.8326,
+      "step": 480
+    },
+    {
+      "epoch": 0.18866108383867553,
+      "grad_norm": 0.24006062746047974,
+      "learning_rate": 9.35740512236047e-06,
+      "loss": 0.8182,
+      "step": 490
+    },
+    {
+      "epoch": 0.1925113100394648,
+      "grad_norm": 0.23032981157302856,
+      "learning_rate": 9.326484890517809e-06,
+      "loss": 0.8227,
+      "step": 500
+    },
+    {
+      "epoch": 0.19636153624025413,
+      "grad_norm": 0.23919883370399475,
+      "learning_rate": 9.294891723898926e-06,
+      "loss": 0.8194,
+      "step": 510
+    },
+    {
+      "epoch": 0.2002117624410434,
+      "grad_norm": 0.26390329003334045,
+      "learning_rate": 9.262630536456153e-06,
+      "loss": 0.8489,
+      "step": 520
+    },
+    {
+      "epoch": 0.2040619886418327,
+      "grad_norm": 0.2786611318588257,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.82,
+      "step": 530
+    },
+    {
+      "epoch": 0.207912214842622,
+      "grad_norm": 0.26839500665664673,
+      "learning_rate": 9.196124273642422e-06,
+      "loss": 0.8342,
+      "step": 540
+    },
+    {
+      "epoch": 0.2117624410434113,
+      "grad_norm": 0.2339692860841751,
+      "learning_rate": 9.161889542552842e-06,
+      "loss": 0.8354,
+      "step": 550
+    },
+    {
+      "epoch": 0.21561266724420058,
+      "grad_norm": 0.24756456911563873,
+      "learning_rate": 9.127007477593193e-06,
+      "loss": 0.8414,
+      "step": 560
+    },
+    {
+      "epoch": 0.2194628934449899,
+      "grad_norm": 0.27925366163253784,
+      "learning_rate": 9.091483504265974e-06,
+      "loss": 0.8509,
+      "step": 570
+    },
+    {
+      "epoch": 0.22331311964577918,
+      "grad_norm": 0.26895344257354736,
+      "learning_rate": 9.055323147915116e-06,
+      "loss": 0.8295,
+      "step": 580
+    },
+    {
+      "epoch": 0.2271633458465685,
+      "grad_norm": 0.2665526866912842,
+      "learning_rate": 9.018532032866585e-06,
+      "loss": 0.8399,
+      "step": 590
+    },
+    {
+      "epoch": 0.23101357204735778,
+      "grad_norm": 0.25341087579727173,
+      "learning_rate": 8.981115881553579e-06,
+      "loss": 0.831,
+      "step": 600
+    },
+    {
+      "epoch": 0.23486379824814707,
+      "grad_norm": 0.27193889021873474,
+      "learning_rate": 8.943080513626476e-06,
+      "loss": 0.8507,
+      "step": 610
+    },
+    {
+      "epoch": 0.23871402444893638,
+      "grad_norm": 0.2525129020214081,
+      "learning_rate": 8.904431845047641e-06,
+      "loss": 0.8343,
+      "step": 620
+    },
+    {
+      "epoch": 0.24256425064972567,
+      "grad_norm": 0.25718119740486145,
+      "learning_rate": 8.86517588717128e-06,
+      "loss": 0.8517,
+      "step": 630
+    },
+    {
+      "epoch": 0.24641447685051496,
+      "grad_norm": 0.24111559987068176,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.8401,
+      "step": 640
+    },
+    {
+      "epoch": 0.25026470305130427,
+      "grad_norm": 0.23562397062778473,
+      "learning_rate": 8.784866620277314e-06,
+      "loss": 0.7969,
+      "step": 650
+    },
+    {
+      "epoch": 0.25411492925209356,
+      "grad_norm": 0.282107412815094,
+      "learning_rate": 8.743825802439018e-06,
+      "loss": 0.8348,
+      "step": 660
+    },
+    {
+      "epoch": 0.25796515545288284,
+      "grad_norm": 0.29429343342781067,
+      "learning_rate": 8.702202675718956e-06,
+      "loss": 0.8394,
+      "step": 670
+    },
+    {
+      "epoch": 0.26181538165367213,
+      "grad_norm": 0.2583483159542084,
+      "learning_rate": 8.660003714113959e-06,
+      "loss": 0.8102,
+      "step": 680
+    },
+    {
+      "epoch": 0.26566560785446147,
+      "grad_norm": 0.24902614951133728,
+      "learning_rate": 8.617235481185323e-06,
+      "loss": 0.835,
+      "step": 690
+    },
+    {
+      "epoch": 0.26951583405525076,
+      "grad_norm": 0.2536700963973999,
+      "learning_rate": 8.573904629037931e-06,
+      "loss": 0.8388,
+      "step": 700
+    },
+    {
+      "epoch": 0.27336606025604004,
+      "grad_norm": 0.24473930895328522,
+      "learning_rate": 8.530017897285585e-06,
+      "loss": 0.819,
+      "step": 710
+    },
+    {
+      "epoch": 0.27721628645682933,
+      "grad_norm": 0.23735162615776062,
+      "learning_rate": 8.485582112002746e-06,
+      "loss": 0.823,
+      "step": 720
+    },
+    {
+      "epoch": 0.2810665126576186,
+      "grad_norm": 0.25260043144226074,
+      "learning_rate": 8.440604184662811e-06,
+      "loss": 0.8293,
+      "step": 730
+    },
+    {
+      "epoch": 0.28491673885840796,
+      "grad_norm": 0.2608690857887268,
+      "learning_rate": 8.395091111063118e-06,
+      "loss": 0.8184,
+      "step": 740
+    },
+    {
+      "epoch": 0.28876696505919724,
+      "grad_norm": 0.23778389394283295,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.8172,
+      "step": 750
+    },
+    {
+      "epoch": 0.29261719125998653,
+      "grad_norm": 0.24516651034355164,
+      "learning_rate": 8.302487923351842e-06,
+      "loss": 0.8101,
+      "step": 760
+    },
+    {
+      "epoch": 0.2964674174607758,
+      "grad_norm": 0.2641059458255768,
+      "learning_rate": 8.255412212597021e-06,
+      "loss": 0.8271,
+      "step": 770
+    },
+    {
+      "epoch": 0.3003176436615651,
+      "grad_norm": 0.24733397364616394,
+      "learning_rate": 8.207830160055688e-06,
+      "loss": 0.8243,
+      "step": 780
+    },
+    {
+      "epoch": 0.3041678698623544,
+      "grad_norm": 0.2611391544342041,
+      "learning_rate": 8.159749166566786e-06,
+      "loss": 0.8096,
+      "step": 790
+    },
+    {
+      "epoch": 0.30801809606314373,
+      "grad_norm": 0.24357880651950836,
+      "learning_rate": 8.11117671057377e-06,
+      "loss": 0.837,
+      "step": 800
+    },
+    {
+      "epoch": 0.311868322263933,
+      "grad_norm": 0.25574174523353577,
+      "learning_rate": 8.062120346961404e-06,
+      "loss": 0.8291,
+      "step": 810
+    },
+    {
+      "epoch": 0.3157185484647223,
+      "grad_norm": 0.25519952178001404,
+      "learning_rate": 8.012587705880705e-06,
+      "loss": 0.8215,
+      "step": 820
+    },
+    {
+      "epoch": 0.3195687746655116,
+      "grad_norm": 0.2525583505630493,
+      "learning_rate": 7.962586491562146e-06,
+      "loss": 0.8083,
+      "step": 830
+    },
+    {
+      "epoch": 0.3234190008663009,
+      "grad_norm": 0.25792673230171204,
+      "learning_rate": 7.912124481117357e-06,
+      "loss": 0.8141,
+      "step": 840
+    },
+    {
+      "epoch": 0.3272692270670902,
+      "grad_norm": 0.24645738303661346,
+      "learning_rate": 7.861209523329494e-06,
+      "loss": 0.8236,
+      "step": 850
+    },
+    {
+      "epoch": 0.3311194532678795,
+      "grad_norm": 0.26956814527511597,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.8073,
+      "step": 860
+    },
+    {
+      "epoch": 0.3349696794686688,
+      "grad_norm": 0.25839927792549133,
+      "learning_rate": 7.758052511879035e-06,
+      "loss": 0.8186,
+      "step": 870
+    },
+    {
+      "epoch": 0.3388199056694581,
+      "grad_norm": 0.22577954828739166,
+      "learning_rate": 7.705826503098642e-06,
+      "loss": 0.8126,
+      "step": 880
+    },
+    {
+      "epoch": 0.34267013187024736,
+      "grad_norm": 0.2461995631456375,
+      "learning_rate": 7.653179634243979e-06,
+      "loss": 0.8124,
+      "step": 890
+    },
+    {
+      "epoch": 0.34652035807103665,
+      "grad_norm": 0.23366709053516388,
+      "learning_rate": 7.600120093927696e-06,
+      "loss": 0.8198,
+      "step": 900
+    },
+    {
+      "epoch": 0.350370584271826,
+      "grad_norm": 0.25407612323760986,
+      "learning_rate": 7.546656134948729e-06,
+      "loss": 0.8092,
+      "step": 910
+    },
+    {
+      "epoch": 0.3542208104726153,
+      "grad_norm": 0.25640979409217834,
+      "learning_rate": 7.492796073008661e-06,
+      "loss": 0.8063,
+      "step": 920
+    },
+    {
+      "epoch": 0.35807103667340456,
+      "grad_norm": 0.26908519864082336,
+      "learning_rate": 7.438548285418326e-06,
+      "loss": 0.7929,
+      "step": 930
+    },
+    {
+      "epoch": 0.36192126287419385,
+      "grad_norm": 0.2643343210220337,
+      "learning_rate": 7.3839212097948e-06,
+      "loss": 0.826,
+      "step": 940
+    },
+    {
+      "epoch": 0.36577148907498314,
+      "grad_norm": 0.26001086831092834,
+      "learning_rate": 7.328923342749039e-06,
+      "loss": 0.8421,
+      "step": 950
+    },
+    {
+      "epoch": 0.3696217152757725,
+      "grad_norm": 0.2657836079597473,
+      "learning_rate": 7.273563238564319e-06,
+      "loss": 0.8256,
+      "step": 960
+    },
+    {
+      "epoch": 0.37347194147656176,
+      "grad_norm": 0.2573256492614746,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.8076,
+      "step": 970
+    },
+    {
+      "epoch": 0.37732216767735105,
+      "grad_norm": 0.2434735745191574,
+      "learning_rate": 7.161790816280858e-06,
+      "loss": 0.8112,
+      "step": 980
+    },
+    {
+      "epoch": 0.38117239387814034,
+      "grad_norm": 0.27291667461395264,
+      "learning_rate": 7.1053958830920125e-06,
+      "loss": 0.799,
+      "step": 990
+    },
+    {
+      "epoch": 0.3850226200789296,
+      "grad_norm": 0.25538283586502075,
+      "learning_rate": 7.0486734798799715e-06,
+      "loss": 0.814,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3888728462797189,
+      "grad_norm": 0.25588133931159973,
+      "learning_rate": 6.991632429159708e-06,
+      "loss": 0.837,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39272307248050825,
+      "grad_norm": 0.25828173756599426,
+      "learning_rate": 6.934281603008132e-06,
+      "loss": 0.8303,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39657329868129754,
+      "grad_norm": 0.2482399195432663,
+      "learning_rate": 6.876629921684143e-06,
+      "loss": 0.8267,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4004235248820868,
+      "grad_norm": 0.2684035897254944,
+      "learning_rate": 6.818686352241199e-06,
+      "loss": 0.8178,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4042737510828761,
+      "grad_norm": 0.2255905717611313,
+      "learning_rate": 6.760459907132573e-06,
+      "loss": 0.8137,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4081239772836654,
+      "grad_norm": 0.23208127915859222,
+      "learning_rate": 6.701959642809589e-06,
+      "loss": 0.8203,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41197420348445474,
+      "grad_norm": 0.27280011773109436,
+      "learning_rate": 6.643194658312987e-06,
+      "loss": 0.8129,
+      "step": 1070
+    },
+    {
+      "epoch": 0.415824429685244,
+      "grad_norm": 0.23379339277744293,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.8363,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4196746558860333,
+      "grad_norm": 0.26759445667266846,
+      "learning_rate": 6.5249071294110735e-06,
+      "loss": 0.8161,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4235248820868226,
+      "grad_norm": 0.26602357625961304,
+      "learning_rate": 6.465402983265274e-06,
+      "loss": 0.8309,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4273751082876119,
+      "grad_norm": 0.2573372721672058,
+      "learning_rate": 6.405670910603243e-06,
+      "loss": 0.8268,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43122533448840117,
+      "grad_norm": 0.2780662775039673,
+      "learning_rate": 6.345720202059291e-06,
+      "loss": 0.8222,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4350755606891905,
+      "grad_norm": 0.2563292384147644,
+      "learning_rate": 6.285560182274008e-06,
+      "loss": 0.8013,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4389257868899798,
+      "grad_norm": 0.24554631114006042,
+      "learning_rate": 6.225200208443936e-06,
+      "loss": 0.8239,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4427760130907691,
+      "grad_norm": 0.23631788790225983,
+      "learning_rate": 6.16464966886616e-06,
+      "loss": 0.8144,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44662623929155837,
+      "grad_norm": 0.27572935819625854,
+      "learning_rate": 6.103917981478065e-06,
+      "loss": 0.8367,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45047646549234766,
+      "grad_norm": 0.26254040002822876,
+      "learning_rate": 6.043014592392489e-06,
+      "loss": 0.8488,
+      "step": 1170
+    },
+    {
+      "epoch": 0.454326691693137,
+      "grad_norm": 0.24071668088436127,
+      "learning_rate": 5.981948974428485e-06,
+      "loss": 0.8225,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4581769178939263,
+      "grad_norm": 0.26665475964546204,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.803,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46202714409471557,
+      "grad_norm": 0.26732775568962097,
+      "learning_rate": 5.8593690678282355e-06,
+      "loss": 0.8121,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46587737029550486,
+      "grad_norm": 0.26380813121795654,
+      "learning_rate": 5.797873845081301e-06,
+      "loss": 0.8162,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46972759649629414,
+      "grad_norm": 0.2500866949558258,
+      "learning_rate": 5.736254522269075e-06,
+      "loss": 0.8335,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47357782269708343,
+      "grad_norm": 0.248367577791214,
+      "learning_rate": 5.6745206835658386e-06,
+      "loss": 0.8339,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47742804889787277,
+      "grad_norm": 0.2475503534078598,
+      "learning_rate": 5.612681930957495e-06,
+      "loss": 0.8049,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48127827509866206,
+      "grad_norm": 0.2391473799943924,
+      "learning_rate": 5.550747882748095e-06,
+      "loss": 0.8042,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48512850129945134,
+      "grad_norm": 0.27549779415130615,
+      "learning_rate": 5.4887281720638195e-06,
+      "loss": 0.8093,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48897872750024063,
+      "grad_norm": 0.2704520523548126,
+      "learning_rate": 5.42663244535466e-06,
+      "loss": 0.7978,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4928289537010299,
+      "grad_norm": 0.2665030360221863,
+      "learning_rate": 5.3644703608940205e-06,
+      "loss": 0.8031,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49667917990181926,
+      "grad_norm": 0.2528025507926941,
+      "learning_rate": 5.302251587276491e-06,
+      "loss": 0.8153,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5005294061026085,
+      "grad_norm": 0.2431778758764267,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.8148,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5043796323033978,
+      "grad_norm": 0.24854432046413422,
+      "learning_rate": 5.177682689530615e-06,
+      "loss": 0.829,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5082298585041871,
+      "grad_norm": 0.2562977969646454,
+      "learning_rate": 5.11535194065619e-06,
+      "loss": 0.7972,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5120800847049765,
+      "grad_norm": 0.2727588415145874,
+      "learning_rate": 5.053003250119113e-06,
+      "loss": 0.809,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5159303109057657,
+      "grad_norm": 0.26171502470970154,
+      "learning_rate": 4.9906463155383925e-06,
+      "loss": 0.8153,
+      "step": 1340
+    },
+    {
+      "epoch": 0.519780537106555,
+      "grad_norm": 0.24831360578536987,
+      "learning_rate": 4.9282908358153046e-06,
+      "loss": 0.806,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5236307633073443,
+      "grad_norm": 0.2855852246284485,
+      "learning_rate": 4.8659465096248345e-06,
+      "loss": 0.8256,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5274809895081336,
+      "grad_norm": 0.2724645733833313,
+      "learning_rate": 4.803623033907169e-06,
+      "loss": 0.8253,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5313312157089229,
+      "grad_norm": 0.23590479791164398,
+      "learning_rate": 4.741330102359441e-06,
+      "loss": 0.829,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5351814419097122,
+      "grad_norm": 0.2593459486961365,
+      "learning_rate": 4.679077403927996e-06,
+      "loss": 0.8099,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5390316681105015,
+      "grad_norm": 0.24805566668510437,
+      "learning_rate": 4.616874621301373e-06,
+      "loss": 0.8082,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5428818943112907,
+      "grad_norm": 0.26098397374153137,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.8237,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5467321205120801,
+      "grad_norm": 0.2680773437023163,
+      "learning_rate": 4.492657493892822e-06,
+      "loss": 0.81,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5505823467128694,
+      "grad_norm": 0.3028339743614197,
+      "learning_rate": 4.430662469650999e-06,
+      "loss": 0.8243,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5544325729136587,
+      "grad_norm": 0.27837735414505005,
+      "learning_rate": 4.368755999289118e-06,
+      "loss": 0.8251,
+      "step": 1440
+    },
+    {
+      "epoch": 0.558282799114448,
+      "grad_norm": 0.27536246180534363,
+      "learning_rate": 4.306947711643949e-06,
+      "loss": 0.8126,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5621330253152372,
+      "grad_norm": 0.2525521218776703,
+      "learning_rate": 4.245247220281068e-06,
+      "loss": 0.8139,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5659832515160266,
+      "grad_norm": 0.2533118426799774,
+      "learning_rate": 4.183664121999587e-06,
+      "loss": 0.8132,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5698334777168159,
+      "grad_norm": 0.2777596712112427,
+      "learning_rate": 4.122207995339474e-06,
+      "loss": 0.8075,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5736837039176051,
+      "grad_norm": 0.2398650050163269,
+      "learning_rate": 4.060888399091734e-06,
+      "loss": 0.8146,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5775339301183945,
+      "grad_norm": 0.24250467121601105,
+      "learning_rate": 3.999714870811639e-06,
+      "loss": 0.8207,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5813841563191837,
+      "grad_norm": 0.25024741888046265,
+      "learning_rate": 3.938696925335275e-06,
+      "loss": 0.8152,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5852343825199731,
+      "grad_norm": 0.25702235102653503,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.8232,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5890846087207623,
+      "grad_norm": 0.25552645325660706,
+      "learning_rate": 3.817165719666382e-06,
+      "loss": 0.8261,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5929348349215516,
+      "grad_norm": 0.24789029359817505,
+      "learning_rate": 3.756671362249841e-06,
+      "loss": 0.8287,
+      "step": 1540
+    },
+    {
+      "epoch": 0.596785061122341,
+      "grad_norm": 0.2523633539676666,
+      "learning_rate": 3.6963703902488907e-06,
+      "loss": 0.8156,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6006352873231302,
+      "grad_norm": 0.24985861778259277,
+      "learning_rate": 3.6362721827835655e-06,
+      "loss": 0.8362,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6044855135239195,
+      "grad_norm": 0.2653239667415619,
+      "learning_rate": 3.576386087436212e-06,
+      "loss": 0.8282,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6083357397247088,
+      "grad_norm": 0.2712172269821167,
+      "learning_rate": 3.5167214187975913e-06,
+      "loss": 0.818,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6121859659254981,
+      "grad_norm": 0.22515125572681427,
+      "learning_rate": 3.457287457018086e-06,
+      "loss": 0.8198,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6160361921262875,
+      "grad_norm": 0.24347274005413055,
+      "learning_rate": 3.3980934463643003e-06,
+      "loss": 0.8076,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6198864183270767,
+      "grad_norm": 0.22842027246952057,
+      "learning_rate": 3.3391485937812053e-06,
+      "loss": 0.79,
+      "step": 1610
+    },
+    {
+      "epoch": 0.623736644527866,
+      "grad_norm": 0.24475571513175964,
+      "learning_rate": 3.280462067460115e-06,
+      "loss": 0.7898,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6275868707286553,
+      "grad_norm": 0.2479376494884491,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.8196,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6314370969294446,
+      "grad_norm": 0.23030680418014526,
+      "learning_rate": 3.163900464051089e-06,
+      "loss": 0.81,
+      "step": 1640
+    },
+    {
+      "epoch": 0.635287323130234,
+      "grad_norm": 0.25493401288986206,
+      "learning_rate": 3.1060435167748826e-06,
+      "loss": 0.8185,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6391375493310232,
+      "grad_norm": 0.24596074223518372,
+      "learning_rate": 3.048481152564253e-06,
+      "loss": 0.7981,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6429877755318125,
+      "grad_norm": 0.2635379731655121,
+      "learning_rate": 2.9912223245804017e-06,
+      "loss": 0.8205,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6468380017326018,
+      "grad_norm": 0.2593478858470917,
+      "learning_rate": 2.934275938772981e-06,
+      "loss": 0.8191,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6506882279333911,
+      "grad_norm": 0.29257139563560486,
+      "learning_rate": 2.8776508524948676e-06,
+      "loss": 0.8078,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6545384541341804,
+      "grad_norm": 0.22679287195205688,
+      "learning_rate": 2.821355873124503e-06,
+      "loss": 0.8068,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6583886803349697,
+      "grad_norm": 0.2636577785015106,
+      "learning_rate": 2.7653997566960144e-06,
+      "loss": 0.8202,
+      "step": 1710
+    },
+    {
+      "epoch": 0.662238906535759,
+      "grad_norm": 0.26222866773605347,
+      "learning_rate": 2.709791206537309e-06,
+      "loss": 0.81,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6660891327365482,
+      "grad_norm": 0.2584359049797058,
+      "learning_rate": 2.6545388719163747e-06,
+      "loss": 0.8181,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6699393589373376,
+      "grad_norm": 0.2474125772714615,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.8145,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6737895851381268,
+      "grad_norm": 0.25409144163131714,
+      "learning_rate": 2.545137167997007e-06,
+      "loss": 0.8104,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6776398113389162,
+      "grad_norm": 0.23818092048168182,
+      "learning_rate": 2.4910048148705972e-06,
+      "loss": 0.7996,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6814900375397055,
+      "grad_norm": 0.27053365111351013,
+      "learning_rate": 2.4372627069793327e-06,
+      "loss": 0.8416,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6853402637404947,
+      "grad_norm": 0.2555851638317108,
+      "learning_rate": 2.3839192032876486e-06,
+      "loss": 0.829,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6891904899412841,
+      "grad_norm": 0.24852830171585083,
+      "learning_rate": 2.3309826007617047e-06,
+      "loss": 0.8392,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6930407161420733,
+      "grad_norm": 0.25026935338974,
+      "learning_rate": 2.278461133078881e-06,
+      "loss": 0.8029,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6968909423428626,
+      "grad_norm": 0.254432737827301,
+      "learning_rate": 2.2263629693471258e-06,
+      "loss": 0.794,
+      "step": 1810
+    },
+    {
+      "epoch": 0.700741168543652,
+      "grad_norm": 0.24881316721439362,
+      "learning_rate": 2.1746962128343414e-06,
+      "loss": 0.8197,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7045913947444412,
+      "grad_norm": 0.24165278673171997,
+      "learning_rate": 2.1234688997080183e-06,
+      "loss": 0.8068,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7084416209452306,
+      "grad_norm": 0.2599773705005646,
+      "learning_rate": 2.0726889977853053e-06,
+      "loss": 0.8114,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7122918471460198,
+      "grad_norm": 0.2701084017753601,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.8239,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7161420733468091,
+      "grad_norm": 0.24744485318660736,
+      "learning_rate": 1.9725029496425814e-06,
+      "loss": 0.8163,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7199922995475985,
+      "grad_norm": 0.2527918815612793,
+      "learning_rate": 1.923112386205727e-06,
+      "loss": 0.8138,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7238425257483877,
+      "grad_norm": 0.2399015575647354,
+      "learning_rate": 1.874200397115079e-06,
+      "loss": 0.8201,
+      "step": 1880
+    },
+    {
+      "epoch": 0.727692751949177,
+      "grad_norm": 0.2531912922859192,
+      "learning_rate": 1.825774590065857e-06,
+      "loss": 0.805,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7315429781499663,
+      "grad_norm": 0.2683999240398407,
+      "learning_rate": 1.7778424971332813e-06,
+      "loss": 0.8114,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7353932043507556,
+      "grad_norm": 0.2518446743488312,
+      "learning_rate": 1.7304115736010407e-06,
+      "loss": 0.8115,
+      "step": 1910
+    },
+    {
+      "epoch": 0.739243430551545,
+      "grad_norm": 0.25003039836883545,
+      "learning_rate": 1.6834891968017108e-06,
+      "loss": 0.8023,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7430936567523342,
+      "grad_norm": 0.2639184594154358,
+      "learning_rate": 1.6370826649692884e-06,
+      "loss": 0.8248,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7469438829531235,
+      "grad_norm": 0.24815523624420166,
+      "learning_rate": 1.5911991961040463e-06,
+      "loss": 0.8296,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7507941091539128,
+      "grad_norm": 0.25488728284835815,
+      "learning_rate": 1.545845926849851e-06,
+      "loss": 0.815,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7546443353547021,
+      "grad_norm": 0.2666793465614319,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.8345,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7584945615554913,
+      "grad_norm": 0.2587490677833557,
+      "learning_rate": 1.4567581203207238e-06,
+      "loss": 0.8217,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7623447877562807,
+      "grad_norm": 0.2614048421382904,
+      "learning_rate": 1.413037439625593e-06,
+      "loss": 0.8223,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76619501395707,
+      "grad_norm": 0.27605435252189636,
+      "learning_rate": 1.3698746695458847e-06,
+      "loss": 0.8143,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7700452401578592,
+      "grad_norm": 0.24868905544281006,
+      "learning_rate": 1.3272765235521745e-06,
+      "loss": 0.8062,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7738954663586486,
+      "grad_norm": 0.23662106692790985,
+      "learning_rate": 1.2852496272942877e-06,
+      "loss": 0.8135,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7777456925594378,
+      "grad_norm": 0.23707003891468048,
+      "learning_rate": 1.2438005175707396e-06,
+      "loss": 0.8029,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7815959187602272,
+      "grad_norm": 0.2396177053451538,
+      "learning_rate": 1.2029356413120297e-06,
+      "loss": 0.7894,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7854461449610165,
+      "grad_norm": 0.2577609419822693,
+      "learning_rate": 1.1626613545778786e-06,
+      "loss": 0.8059,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7892963711618057,
+      "grad_norm": 0.2342158406972885,
+      "learning_rate": 1.1229839215686294e-06,
+      "loss": 0.8101,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7931465973625951,
+      "grad_norm": 0.22986851632595062,
+      "learning_rate": 1.0839095136509188e-06,
+      "loss": 0.7969,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7969968235633843,
+      "grad_norm": 0.2680145800113678,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.8191,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8008470497641736,
+      "grad_norm": 0.24304227530956268,
+      "learning_rate": 1.007593988643402e-06,
+      "loss": 0.7972,
+      "step": 2080
+    },
+    {
+      "epoch": 0.804697275964963,
+      "grad_norm": 0.23707719147205353,
+      "learning_rate": 9.703647415524613e-07,
+      "loss": 0.8192,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8085475021657522,
+      "grad_norm": 0.24732525646686554,
+      "learning_rate": 9.337622577045485e-07,
+      "loss": 0.8088,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8123977283665416,
+      "grad_norm": 0.24396587908267975,
+      "learning_rate": 8.977922301934588e-07,
+      "loss": 0.8158,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8162479545673308,
+      "grad_norm": 0.24798737466335297,
+      "learning_rate": 8.624602537417009e-07,
+      "loss": 0.8179,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8200981807681201,
+      "grad_norm": 0.26980823278427124,
+      "learning_rate": 8.277718238303123e-07,
+      "loss": 0.8076,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8239484069689095,
+      "grad_norm": 0.2541809380054474,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.8102,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8277986331696987,
+      "grad_norm": 0.26155298948287964,
+      "learning_rate": 7.603470842324223e-07,
+      "loss": 0.8031,
+      "step": 2150
+    },
+    {
+      "epoch": 0.831648859370488,
+      "grad_norm": 0.23454512655735016,
+      "learning_rate": 7.276212616857575e-07,
+      "loss": 0.8193,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8354990855712773,
+      "grad_norm": 0.23491176962852478,
+      "learning_rate": 6.955599583279837e-07,
+      "loss": 0.8217,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8393493117720666,
+      "grad_norm": 0.24505455791950226,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.8104,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8431995379728559,
+      "grad_norm": 0.2456839233636856,
+      "learning_rate": 6.334507521075711e-07,
+      "loss": 0.8213,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8470497641736452,
+      "grad_norm": 0.2603183686733246,
+      "learning_rate": 6.034125096149179e-07,
+      "loss": 0.8119,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8508999903744345,
+      "grad_norm": 0.2469349354505539,
+      "learning_rate": 5.740581055485722e-07,
+      "loss": 0.7838,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8547502165752238,
+      "grad_norm": 0.2330351024866104,
+      "learning_rate": 5.453921056471961e-07,
+      "loss": 0.8192,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8586004427760131,
+      "grad_norm": 0.2675358057022095,
+      "learning_rate": 5.174189685761288e-07,
+      "loss": 0.8108,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8624506689768023,
+      "grad_norm": 0.2534945607185364,
+      "learning_rate": 4.901430452338979e-07,
+      "loss": 0.8223,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8663008951775917,
+      "grad_norm": 0.2441825419664383,
+      "learning_rate": 4.6356857807548286e-07,
+      "loss": 0.8073,
+      "step": 2250
+    },
+    {
+      "epoch": 0.870151121378381,
+      "grad_norm": 0.24762341380119324,
+      "learning_rate": 4.376997004524541e-07,
+      "loss": 0.7965,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8740013475791703,
+      "grad_norm": 0.25473153591156006,
+      "learning_rate": 4.12540435970073e-07,
+      "loss": 0.8093,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8778515737799596,
+      "grad_norm": 0.2514944076538086,
+      "learning_rate": 3.880946978614703e-07,
+      "loss": 0.8152,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8817017999807488,
+      "grad_norm": 0.24565820395946503,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.823,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8855520261815382,
+      "grad_norm": 0.2241189032793045,
+      "learning_rate": 3.41358898202781e-07,
+      "loss": 0.7983,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8894022523823275,
+      "grad_norm": 0.23332157731056213,
+      "learning_rate": 3.190761058667741e-07,
+      "loss": 0.8095,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8932524785831167,
+      "grad_norm": 0.2332276850938797,
+      "learning_rate": 2.9752137720206463e-07,
+      "loss": 0.8172,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8971027047839061,
+      "grad_norm": 0.25302714109420776,
+      "learning_rate": 2.7669806479785067e-07,
+      "loss": 0.8207,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9009529309846953,
+      "grad_norm": 0.2475683093070984,
+      "learning_rate": 2.566094074799763e-07,
+      "loss": 0.8166,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9048031571854847,
+      "grad_norm": 0.24248559772968292,
+      "learning_rate": 2.3725852980716557e-07,
+      "loss": 0.7955,
+      "step": 2350
+    },
+    {
+      "epoch": 0.908653383386274,
+      "grad_norm": 0.2435782104730606,
+      "learning_rate": 2.1864844158503983e-07,
+      "loss": 0.8094,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9125036095870632,
+      "grad_norm": 0.278368204832077,
+      "learning_rate": 2.007820373979702e-07,
+      "loss": 0.8013,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9163538357878526,
+      "grad_norm": 0.24541805684566498,
+      "learning_rate": 1.8366209615886532e-07,
+      "loss": 0.8093,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9202040619886418,
+      "grad_norm": 0.24156567454338074,
+      "learning_rate": 1.6729128067693613e-07,
+      "loss": 0.8184,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9240542881894311,
+      "grad_norm": 0.26375964283943176,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.8199,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9279045143902205,
+      "grad_norm": 0.2568744421005249,
+      "learning_rate": 1.3680709523610058e-07,
+      "loss": 0.8038,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9317547405910097,
+      "grad_norm": 0.23997721076011658,
+      "learning_rate": 1.2269846674030418e-07,
+      "loss": 0.8258,
+      "step": 2420
+    },
+    {
+      "epoch": 0.935604966791799,
+      "grad_norm": 0.24411426484584808,
+      "learning_rate": 1.0934844619042595e-07,
+      "loss": 0.8224,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9394551929925883,
+      "grad_norm": 0.2691296935081482,
+      "learning_rate": 9.675911002803961e-08,
+      "loss": 0.8122,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9433054191933776,
+      "grad_norm": 0.2628415524959564,
+      "learning_rate": 8.493241637904381e-08,
+      "loss": 0.8093,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9471556453941669,
+      "grad_norm": 0.24867400527000427,
+      "learning_rate": 7.387020474909689e-08,
+      "loss": 0.8288,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9510058715949562,
+      "grad_norm": 0.2620219886302948,
+      "learning_rate": 6.357419573750856e-08,
+      "loss": 0.8234,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9548560977957455,
+      "grad_norm": 0.24492432177066803,
+      "learning_rate": 5.4045990769613365e-08,
+      "loss": 0.806,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9587063239965348,
+      "grad_norm": 0.2999849319458008,
+      "learning_rate": 4.5287071847694434e-08,
+      "loss": 0.812,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9625565501973241,
+      "grad_norm": 0.24192801117897034,
+      "learning_rate": 3.7298801320467817e-08,
+      "loss": 0.8042,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9664067763981133,
+      "grad_norm": 0.23941992223262787,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.7978,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9702570025989027,
+      "grad_norm": 0.2459937334060669,
+      "learning_rate": 2.3639055324410355e-08,
+      "loss": 0.8195,
+      "step": 2520
+    },
+    {
+      "epoch": 0.974107228799692,
+      "grad_norm": 0.24120256304740906,
+      "learning_rate": 1.7969704471369477e-08,
+      "loss": 0.8043,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9779574550004813,
+      "grad_norm": 0.23698271811008453,
+      "learning_rate": 1.3075250914142413e-08,
+      "loss": 0.8081,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9818076812012706,
+      "grad_norm": 0.23700441420078278,
+      "learning_rate": 8.956455928476049e-09,
+      "loss": 0.8182,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9856579074020598,
+      "grad_norm": 0.2521786093711853,
+      "learning_rate": 5.6139601453786275e-09,
+      "loss": 0.8103,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9895081336028492,
+      "grad_norm": 0.24455921351909637,
+      "learning_rate": 3.0482834514805602e-09,
+      "loss": 0.7975,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9933583598036385,
+      "grad_norm": 0.225899338722229,
+      "learning_rate": 1.2598249081713366e-09,
+      "loss": 0.7936,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9972085860044277,
+      "grad_norm": 0.26654210686683655,
+      "learning_rate": 2.488626895302826e-10,
+      "loss": 0.8217,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9999037443449803,
+      "step": 2597,
+      "total_flos": 1.100095443110042e+19,
+      "train_loss": 0.8399555904386958,
+      "train_runtime": 13487.9304,
+      "train_samples_per_second": 49.296,
+      "train_steps_per_second": 0.193
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2597,
+  "num_input_tokens_seen": 418855352,
+  "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": 1.100095443110042e+19,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3/webssl_mae3b_full2b_224/finetune/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36751d8b38a1dd925ef17cf8c7a42e7e4f20f244499dfde90ada5827f1046d14
+size 6264

continuous/qwen3/webssl_mae3b_full2b_224/pretrain/config.json

@@ -0,0 +1,89 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 40960,
+  "max_window_layers": 28,
+  "mm_hidden_size": 3072,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "flat",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_tunable_parts": "mm_mlp_adapter",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -1,
+  "mm_vision_tower": "hf:/cache/ckpt/download/tokenizer/continuous/webssl/webssl-mae3b-full2b-224",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "pad_token_id": 151643,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "transformers_version": "4.57.6",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": null,
+  "vocab_size": 151936
+}

continuous/qwen3/webssl_mae3b_full2b_224/pretrain/mm_projector.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:464ac2280feacf841c225c4bff90d732016352e424eb5dae67826aad98151bd9
+size 41961528

continuous/qwen3/webssl_mae3b_full2b_224/pretrain/trainer_state.json

@@ -0,0 +1,421 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 545,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01834862385321101,
+      "grad_norm": 6.262796401977539,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 8.3673,
+      "step": 10
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 0.5141330361366272,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 5.1964,
+      "step": 20
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.12343034148216248,
+      "learning_rate": 0.0009999907688176173,
+      "loss": 4.7999,
+      "step": 30
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.13668107986450195,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 4.6976,
+      "step": 40
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.04789091274142265,
+      "learning_rate": 0.0009959345572940772,
+      "loss": 4.6603,
+      "step": 50
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.04488211125135422,
+      "learning_rate": 0.0009911550081155983,
+      "loss": 4.63,
+      "step": 60
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.08248601853847504,
+      "learning_rate": 0.0009845624347566433,
+      "loss": 4.5961,
+      "step": 70
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.10407736152410507,
+      "learning_rate": 0.0009761811727013547,
+      "loss": 4.5835,
+      "step": 80
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.08727174252271652,
+      "learning_rate": 0.0009660421601067666,
+      "loss": 4.5582,
+      "step": 90
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.08224718272686005,
+      "learning_rate": 0.0009541828235992926,
+      "loss": 4.5458,
+      "step": 100
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.1198911964893341,
+      "learning_rate": 0.0009406469401200151,
+      "loss": 4.5251,
+      "step": 110
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.16526120901107788,
+      "learning_rate": 0.0009254844753287492,
+      "loss": 4.5204,
+      "step": 120
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.1332959532737732,
+      "learning_rate": 0.0009087513991633924,
+      "loss": 4.4953,
+      "step": 130
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.13665618002414703,
+      "learning_rate": 0.0008905094792353917,
+      "loss": 4.4803,
+      "step": 140
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.11670264601707458,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 4.4789,
+      "step": 150
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.1375475823879242,
+      "learning_rate": 0.0008497737783108237,
+      "loss": 4.4658,
+      "step": 160
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.12511630356311798,
+      "learning_rate": 0.0008274303669726426,
+      "loss": 4.4512,
+      "step": 170
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.10149767249822617,
+      "learning_rate": 0.0008038782961218137,
+      "loss": 4.4398,
+      "step": 180
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.13014566898345947,
+      "learning_rate": 0.000779204504653894,
+      "loss": 4.4367,
+      "step": 190
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.1272425353527069,
+      "learning_rate": 0.0007535000721258791,
+      "loss": 4.4167,
+      "step": 200
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.19143347442150116,
+      "learning_rate": 0.00072685988254984,
+      "loss": 4.4109,
+      "step": 210
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.14306382834911346,
+      "learning_rate": 0.0006993822741429906,
+      "loss": 4.425,
+      "step": 220
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.13560925424098969,
+      "learning_rate": 0.0006711686763270818,
+      "loss": 4.4216,
+      "step": 230
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.14918462932109833,
+      "learning_rate": 0.0006423232353170831,
+      "loss": 4.4053,
+      "step": 240
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.15083761513233185,
+      "learning_rate": 0.0006129524296812335,
+      "loss": 4.3855,
+      "step": 250
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.17883794009685516,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 4.397,
+      "step": 260
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.11258049309253693,
+      "learning_rate": 0.0005530699351157809,
+      "loss": 4.3785,
+      "step": 270
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.12920907139778137,
+      "learning_rate": 0.0005227792933277943,
+      "loss": 4.3757,
+      "step": 280
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.15422578155994415,
+      "learning_rate": 0.0004924045652352092,
+      "loss": 4.3633,
+      "step": 290
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.14619652926921844,
+      "learning_rate": 0.0004620578745374633,
+      "loss": 4.361,
+      "step": 300
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.14385248720645905,
+      "learning_rate": 0.0004318512414382059,
+      "loss": 4.3758,
+      "step": 310
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.10150765627622604,
+      "learning_rate": 0.0004018961691397148,
+      "loss": 4.3648,
+      "step": 320
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.1316201388835907,
+      "learning_rate": 0.0003723032322457458,
+      "loss": 4.3479,
+      "step": 330
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.11908493936061859,
+      "learning_rate": 0.0003431816685921625,
+      "loss": 4.3486,
+      "step": 340
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.11052057892084122,
+      "learning_rate": 0.0003146389760120416,
+      "loss": 4.358,
+      "step": 350
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.12816733121871948,
+      "learning_rate": 0.00028678051552373485,
+      "loss": 4.3605,
+      "step": 360
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.1196424812078476,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 4.3447,
+      "step": 370
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.14027555286884308,
+      "learning_rate": 0.0002335247266004629,
+      "loss": 4.359,
+      "step": 380
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.14043788611888885,
+      "learning_rate": 0.000208323983828828,
+      "loss": 4.3413,
+      "step": 390
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.13189227879047394,
+      "learning_rate": 0.0001841999188095224,
+      "loss": 4.3447,
+      "step": 400
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.09708092361688614,
+      "learning_rate": 0.00016124158186777676,
+      "loss": 4.343,
+      "step": 410
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.11841386556625366,
+      "learning_rate": 0.0001395337202205211,
+      "loss": 4.3341,
+      "step": 420
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.09252115339040756,
+      "learning_rate": 0.00011915646514489292,
+      "loss": 4.3418,
+      "step": 430
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.1145525574684143,
+      "learning_rate": 0.00010018503618578818,
+      "loss": 4.3475,
+      "step": 440
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.12189189344644547,
+      "learning_rate": 8.268946349432583e-05,
+      "loss": 4.3378,
+      "step": 450
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.12028723210096359,
+      "learning_rate": 6.673432932217117e-05,
+      "loss": 4.3314,
+      "step": 460
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.12409501522779465,
+      "learning_rate": 5.2378529625954684e-05,
+      "loss": 4.3169,
+      "step": 470
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.09091485291719437,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 4.3267,
+      "step": 480
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.08517304062843323,
+      "learning_rate": 2.8670803372379538e-05,
+      "loss": 4.3299,
+      "step": 490
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.08662626147270203,
+      "learning_rate": 1.9406390288876586e-05,
+      "loss": 4.3428,
+      "step": 500
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.08460429310798645,
+      "learning_rate": 1.191601558631461e-05,
+      "loss": 4.3302,
+      "step": 510
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.07644754648208618,
+      "learning_rate": 6.2273288462559305e-06,
+      "loss": 4.3216,
+      "step": 520
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.08274001628160477,
+      "learning_rate": 2.361328992552314e-06,
+      "loss": 4.3256,
+      "step": 530
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.09091539680957794,
+      "learning_rate": 3.322867770044202e-07,
+      "loss": 4.3318,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "step": 545,
+      "total_flos": 8.377413038893957e+17,
+      "train_loss": 4.507388218171006,
+      "train_runtime": 5058.4272,
+      "train_samples_per_second": 110.336,
+      "train_steps_per_second": 0.108
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 545,
+  "num_input_tokens_seen": 31896544,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.377413038893957e+17,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3base/mc2_l14_224/finetune/added_tokens.json

@@ -0,0 +1,28 @@
+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 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
+}

continuous/qwen3base/mc2_l14_224/finetune/chat_template.jinja

@@ -0,0 +1,85 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\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 <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set content = message.content %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in message.content %}
+                {%- set content = message.content.split('</think>')[-1].lstrip('\n') %}
+                {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

continuous/qwen3base/mc2_l14_224/finetune/config.json

@@ -0,0 +1,89 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1024,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "flat",
+  "mm_projector_lr": 0.0001,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_tunable_parts": "mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "open_clip_hub:ViT-L-14",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "pad_token_id": 151643,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "transformers_version": "4.57.6",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": "/cache/ckpt/download/tokenizer/continuous/mc2/mc2_l14_224.pt",
+  "vocab_size": 151936
+}

continuous/qwen3base/mc2_l14_224/finetune/generation_config.json

@@ -0,0 +1,8 @@
+{
+  "eos_token_id": [
+    151643
+  ],
+  "max_new_tokens": 2048,
+  "pad_token_id": 151643,
+  "transformers_version": "4.57.6"
+}

continuous/qwen3base/mc2_l14_224/finetune/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c44863b839980b3cbc634e725cf4ee4fdc15b63d8f2d8bdadc63f97d4345cfa
+size 4696673872

continuous/qwen3base/mc2_l14_224/finetune/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": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

continuous/qwen3base/mc2_l14_224/finetune/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
+size 11422654

continuous/qwen3base/mc2_l14_224/finetune/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": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "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": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "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": "<|endoftext|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

continuous/qwen3base/mc2_l14_224/finetune/trainer_state.json

@@ -0,0 +1,1856 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999037443449803,
+  "eval_steps": 500,
+  "global_step": 2597,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003850226200789296,
+      "grad_norm": 5.628152370452881,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 2.9461,
+      "step": 10
+    },
+    {
+      "epoch": 0.007700452401578592,
+      "grad_norm": 4.113055229187012,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 2.7229,
+      "step": 20
+    },
+    {
+      "epoch": 0.01155067860236789,
+      "grad_norm": 3.4043803215026855,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 2.3982,
+      "step": 30
+    },
+    {
+      "epoch": 0.015400904803157185,
+      "grad_norm": 3.051149606704712,
+      "learning_rate": 5e-06,
+      "loss": 2.3019,
+      "step": 40
+    },
+    {
+      "epoch": 0.019251131003946483,
+      "grad_norm": 2.883314371109009,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 2.1404,
+      "step": 50
+    },
+    {
+      "epoch": 0.02310135720473578,
+      "grad_norm": 3.374079942703247,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 2.0484,
+      "step": 60
+    },
+    {
+      "epoch": 0.026951583405525073,
+      "grad_norm": 3.3394217491149902,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 1.9558,
+      "step": 70
+    },
+    {
+      "epoch": 0.03080180960631437,
+      "grad_norm": 3.6184632778167725,
+      "learning_rate": 9.999996111488723e-06,
+      "loss": 1.9046,
+      "step": 80
+    },
+    {
+      "epoch": 0.034652035807103666,
+      "grad_norm": 3.2538464069366455,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 1.7862,
+      "step": 90
+    },
+    {
+      "epoch": 0.038502262007892966,
+      "grad_norm": 3.585264205932617,
+      "learning_rate": 9.998285264324193e-06,
+      "loss": 1.6766,
+      "step": 100
+    },
+    {
+      "epoch": 0.04235248820868226,
+      "grad_norm": 3.52317476272583,
+      "learning_rate": 9.996263605626052e-06,
+      "loss": 1.5649,
+      "step": 110
+    },
+    {
+      "epoch": 0.04620271440947156,
+      "grad_norm": 3.4283149242401123,
+      "learning_rate": 9.99346483580503e-06,
+      "loss": 1.4691,
+      "step": 120
+    },
+    {
+      "epoch": 0.05005294061026085,
+      "grad_norm": 3.024766683578491,
+      "learning_rate": 9.989889390177467e-06,
+      "loss": 1.3617,
+      "step": 130
+    },
+    {
+      "epoch": 0.053903166811050146,
+      "grad_norm": 2.657637596130371,
+      "learning_rate": 9.985537824862647e-06,
+      "loss": 1.2538,
+      "step": 140
+    },
+    {
+      "epoch": 0.057753393011839446,
+      "grad_norm": 2.412651538848877,
+      "learning_rate": 9.980410816696317e-06,
+      "loss": 1.1563,
+      "step": 150
+    },
+    {
+      "epoch": 0.06160361921262874,
+      "grad_norm": 2.1879279613494873,
+      "learning_rate": 9.97450916312541e-06,
+      "loss": 1.078,
+      "step": 160
+    },
+    {
+      "epoch": 0.06545384541341803,
+      "grad_norm": 1.6553536653518677,
+      "learning_rate": 9.967833782084004e-06,
+      "loss": 1.0213,
+      "step": 170
+    },
+    {
+      "epoch": 0.06930407161420733,
+      "grad_norm": 1.3068230152130127,
+      "learning_rate": 9.960385711850554e-06,
+      "loss": 0.9956,
+      "step": 180
+    },
+    {
+      "epoch": 0.07315429781499663,
+      "grad_norm": 1.08774995803833,
+      "learning_rate": 9.952166110886394e-06,
+      "loss": 0.9569,
+      "step": 190
+    },
+    {
+      "epoch": 0.07700452401578593,
+      "grad_norm": 0.852286159992218,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.9253,
+      "step": 200
+    },
+    {
+      "epoch": 0.08085475021657522,
+      "grad_norm": 0.7052011489868164,
+      "learning_rate": 9.933417550425952e-06,
+      "loss": 0.9024,
+      "step": 210
+    },
+    {
+      "epoch": 0.08470497641736452,
+      "grad_norm": 0.8388557434082031,
+      "learning_rate": 9.9228915070518e-06,
+      "loss": 0.8835,
+      "step": 220
+    },
+    {
+      "epoch": 0.08855520261815382,
+      "grad_norm": 0.7808427214622498,
+      "learning_rate": 9.911599764737632e-06,
+      "loss": 0.8617,
+      "step": 230
+    },
+    {
+      "epoch": 0.09240542881894312,
+      "grad_norm": 0.8041413426399231,
+      "learning_rate": 9.899544079783606e-06,
+      "loss": 0.8384,
+      "step": 240
+    },
+    {
+      "epoch": 0.0962556550197324,
+      "grad_norm": 0.7632434368133545,
+      "learning_rate": 9.886726327312338e-06,
+      "loss": 0.8403,
+      "step": 250
+    },
+    {
+      "epoch": 0.1001058812205217,
+      "grad_norm": 0.6472905278205872,
+      "learning_rate": 9.87314850097724e-06,
+      "loss": 0.8212,
+      "step": 260
+    },
+    {
+      "epoch": 0.103956107421311,
+      "grad_norm": 0.5975033640861511,
+      "learning_rate": 9.858812712652452e-06,
+      "loss": 0.7991,
+      "step": 270
+    },
+    {
+      "epoch": 0.10780633362210029,
+      "grad_norm": 0.5304135084152222,
+      "learning_rate": 9.843721192104341e-06,
+      "loss": 0.7895,
+      "step": 280
+    },
+    {
+      "epoch": 0.11165655982288959,
+      "grad_norm": 0.5015220046043396,
+      "learning_rate": 9.827876286644698e-06,
+      "loss": 0.803,
+      "step": 290
+    },
+    {
+      "epoch": 0.11550678602367889,
+      "grad_norm": 0.4906582534313202,
+      "learning_rate": 9.811280460765643e-06,
+      "loss": 0.7657,
+      "step": 300
+    },
+    {
+      "epoch": 0.11935701222446819,
+      "grad_norm": 0.45002859830856323,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.7718,
+      "step": 310
+    },
+    {
+      "epoch": 0.12320723842525748,
+      "grad_norm": 0.4221496880054474,
+      "learning_rate": 9.775846489301279e-06,
+      "loss": 0.7691,
+      "step": 320
+    },
+    {
+      "epoch": 0.12705746462604678,
+      "grad_norm": 0.4759712219238281,
+      "learning_rate": 9.757013855061157e-06,
+      "loss": 0.768,
+      "step": 330
+    },
+    {
+      "epoch": 0.13090769082683607,
+      "grad_norm": 0.43794044852256775,
+      "learning_rate": 9.737441322234758e-06,
+      "loss": 0.7778,
+      "step": 340
+    },
+    {
+      "epoch": 0.13475791702762538,
+      "grad_norm": 0.39192795753479004,
+      "learning_rate": 9.717131935103605e-06,
+      "loss": 0.7764,
+      "step": 350
+    },
+    {
+      "epoch": 0.13860814322841467,
+      "grad_norm": 0.39301958680152893,
+      "learning_rate": 9.696088852558398e-06,
+      "loss": 0.7632,
+      "step": 360
+    },
+    {
+      "epoch": 0.14245836942920398,
+      "grad_norm": 0.38342851400375366,
+      "learning_rate": 9.674315347607686e-06,
+      "loss": 0.7871,
+      "step": 370
+    },
+    {
+      "epoch": 0.14630859562999327,
+      "grad_norm": 0.36520302295684814,
+      "learning_rate": 9.651814806868796e-06,
+      "loss": 0.751,
+      "step": 380
+    },
+    {
+      "epoch": 0.15015882183078255,
+      "grad_norm": 0.3967287838459015,
+      "learning_rate": 9.62859073004107e-06,
+      "loss": 0.7554,
+      "step": 390
+    },
+    {
+      "epoch": 0.15400904803157187,
+      "grad_norm": 0.3545922040939331,
+      "learning_rate": 9.604646729361547e-06,
+      "loss": 0.7564,
+      "step": 400
+    },
+    {
+      "epoch": 0.15785927423236115,
+      "grad_norm": 0.3354630172252655,
+      "learning_rate": 9.579986529043099e-06,
+      "loss": 0.7447,
+      "step": 410
+    },
+    {
+      "epoch": 0.16170950043315044,
+      "grad_norm": 0.36285415291786194,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.7625,
+      "step": 420
+    },
+    {
+      "epoch": 0.16555972663393975,
+      "grad_norm": 0.34369802474975586,
+      "learning_rate": 9.528532982727281e-06,
+      "loss": 0.7528,
+      "step": 430
+    },
+    {
+      "epoch": 0.16940995283472904,
+      "grad_norm": 0.3476774990558624,
+      "learning_rate": 9.501747639735018e-06,
+      "loss": 0.7488,
+      "step": 440
+    },
+    {
+      "epoch": 0.17326017903551832,
+      "grad_norm": 0.361995667219162,
+      "learning_rate": 9.474262101869272e-06,
+      "loss": 0.7402,
+      "step": 450
+    },
+    {
+      "epoch": 0.17711040523630764,
+      "grad_norm": 0.35276883840560913,
+      "learning_rate": 9.446080644188141e-06,
+      "loss": 0.7452,
+      "step": 460
+    },
+    {
+      "epoch": 0.18096063143709692,
+      "grad_norm": 0.3167298138141632,
+      "learning_rate": 9.417207649992017e-06,
+      "loss": 0.7505,
+      "step": 470
+    },
+    {
+      "epoch": 0.18481085763788624,
+      "grad_norm": 0.31858089566230774,
+      "learning_rate": 9.387647610141816e-06,
+      "loss": 0.749,
+      "step": 480
+    },
+    {
+      "epoch": 0.18866108383867553,
+      "grad_norm": 0.31625983119010925,
+      "learning_rate": 9.35740512236047e-06,
+      "loss": 0.7369,
+      "step": 490
+    },
+    {
+      "epoch": 0.1925113100394648,
+      "grad_norm": 0.3139210343360901,
+      "learning_rate": 9.326484890517809e-06,
+      "loss": 0.7443,
+      "step": 500
+    },
+    {
+      "epoch": 0.19636153624025413,
+      "grad_norm": 0.3365113437175751,
+      "learning_rate": 9.294891723898926e-06,
+      "loss": 0.7388,
+      "step": 510
+    },
+    {
+      "epoch": 0.2002117624410434,
+      "grad_norm": 0.34522274136543274,
+      "learning_rate": 9.262630536456153e-06,
+      "loss": 0.7659,
+      "step": 520
+    },
+    {
+      "epoch": 0.2040619886418327,
+      "grad_norm": 0.3373745381832123,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.7376,
+      "step": 530
+    },
+    {
+      "epoch": 0.207912214842622,
+      "grad_norm": 0.33587023615837097,
+      "learning_rate": 9.196124273642422e-06,
+      "loss": 0.7509,
+      "step": 540
+    },
+    {
+      "epoch": 0.2117624410434113,
+      "grad_norm": 0.29344967007637024,
+      "learning_rate": 9.161889542552842e-06,
+      "loss": 0.7557,
+      "step": 550
+    },
+    {
+      "epoch": 0.21561266724420058,
+      "grad_norm": 0.3233027756214142,
+      "learning_rate": 9.127007477593193e-06,
+      "loss": 0.7578,
+      "step": 560
+    },
+    {
+      "epoch": 0.2194628934449899,
+      "grad_norm": 0.44792360067367554,
+      "learning_rate": 9.091483504265974e-06,
+      "loss": 0.7614,
+      "step": 570
+    },
+    {
+      "epoch": 0.22331311964577918,
+      "grad_norm": 0.321279913187027,
+      "learning_rate": 9.055323147915116e-06,
+      "loss": 0.745,
+      "step": 580
+    },
+    {
+      "epoch": 0.2271633458465685,
+      "grad_norm": 0.3126884400844574,
+      "learning_rate": 9.018532032866585e-06,
+      "loss": 0.7543,
+      "step": 590
+    },
+    {
+      "epoch": 0.23101357204735778,
+      "grad_norm": 0.30662667751312256,
+      "learning_rate": 8.981115881553579e-06,
+      "loss": 0.7467,
+      "step": 600
+    },
+    {
+      "epoch": 0.23486379824814707,
+      "grad_norm": 0.34248998761177063,
+      "learning_rate": 8.943080513626476e-06,
+      "loss": 0.7651,
+      "step": 610
+    },
+    {
+      "epoch": 0.23871402444893638,
+      "grad_norm": 0.3015410602092743,
+      "learning_rate": 8.904431845047641e-06,
+      "loss": 0.7505,
+      "step": 620
+    },
+    {
+      "epoch": 0.24256425064972567,
+      "grad_norm": 0.30771905183792114,
+      "learning_rate": 8.86517588717128e-06,
+      "loss": 0.7679,
+      "step": 630
+    },
+    {
+      "epoch": 0.24641447685051496,
+      "grad_norm": 0.3066297471523285,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.7567,
+      "step": 640
+    },
+    {
+      "epoch": 0.25026470305130427,
+      "grad_norm": 0.3196113705635071,
+      "learning_rate": 8.784866620277314e-06,
+      "loss": 0.7164,
+      "step": 650
+    },
+    {
+      "epoch": 0.25411492925209356,
+      "grad_norm": 0.3267951011657715,
+      "learning_rate": 8.743825802439018e-06,
+      "loss": 0.7472,
+      "step": 660
+    },
+    {
+      "epoch": 0.25796515545288284,
+      "grad_norm": 0.3210297226905823,
+      "learning_rate": 8.702202675718956e-06,
+      "loss": 0.7571,
+      "step": 670
+    },
+    {
+      "epoch": 0.26181538165367213,
+      "grad_norm": 0.29513248801231384,
+      "learning_rate": 8.660003714113959e-06,
+      "loss": 0.7246,
+      "step": 680
+    },
+    {
+      "epoch": 0.26566560785446147,
+      "grad_norm": 0.297048419713974,
+      "learning_rate": 8.617235481185323e-06,
+      "loss": 0.7477,
+      "step": 690
+    },
+    {
+      "epoch": 0.26951583405525076,
+      "grad_norm": 0.3051026165485382,
+      "learning_rate": 8.573904629037931e-06,
+      "loss": 0.755,
+      "step": 700
+    },
+    {
+      "epoch": 0.27336606025604004,
+      "grad_norm": 0.30137747526168823,
+      "learning_rate": 8.530017897285585e-06,
+      "loss": 0.7365,
+      "step": 710
+    },
+    {
+      "epoch": 0.27721628645682933,
+      "grad_norm": 0.28347864747047424,
+      "learning_rate": 8.485582112002746e-06,
+      "loss": 0.7409,
+      "step": 720
+    },
+    {
+      "epoch": 0.2810665126576186,
+      "grad_norm": 0.292850136756897,
+      "learning_rate": 8.440604184662811e-06,
+      "loss": 0.7426,
+      "step": 730
+    },
+    {
+      "epoch": 0.28491673885840796,
+      "grad_norm": 0.3018176555633545,
+      "learning_rate": 8.395091111063118e-06,
+      "loss": 0.731,
+      "step": 740
+    },
+    {
+      "epoch": 0.28876696505919724,
+      "grad_norm": 0.28120157122612,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.7332,
+      "step": 750
+    },
+    {
+      "epoch": 0.29261719125998653,
+      "grad_norm": 0.28579041361808777,
+      "learning_rate": 8.302487923351842e-06,
+      "loss": 0.724,
+      "step": 760
+    },
+    {
+      "epoch": 0.2964674174607758,
+      "grad_norm": 0.2863207459449768,
+      "learning_rate": 8.255412212597021e-06,
+      "loss": 0.7411,
+      "step": 770
+    },
+    {
+      "epoch": 0.3003176436615651,
+      "grad_norm": 0.281093031167984,
+      "learning_rate": 8.207830160055688e-06,
+      "loss": 0.7384,
+      "step": 780
+    },
+    {
+      "epoch": 0.3041678698623544,
+      "grad_norm": 0.3081432282924652,
+      "learning_rate": 8.159749166566786e-06,
+      "loss": 0.7248,
+      "step": 790
+    },
+    {
+      "epoch": 0.30801809606314373,
+      "grad_norm": 0.2801932990550995,
+      "learning_rate": 8.11117671057377e-06,
+      "loss": 0.7501,
+      "step": 800
+    },
+    {
+      "epoch": 0.311868322263933,
+      "grad_norm": 0.30353274941444397,
+      "learning_rate": 8.062120346961404e-06,
+      "loss": 0.743,
+      "step": 810
+    },
+    {
+      "epoch": 0.3157185484647223,
+      "grad_norm": 0.2740929424762726,
+      "learning_rate": 8.012587705880705e-06,
+      "loss": 0.7369,
+      "step": 820
+    },
+    {
+      "epoch": 0.3195687746655116,
+      "grad_norm": 0.30588021874427795,
+      "learning_rate": 7.962586491562146e-06,
+      "loss": 0.723,
+      "step": 830
+    },
+    {
+      "epoch": 0.3234190008663009,
+      "grad_norm": 0.31052008271217346,
+      "learning_rate": 7.912124481117357e-06,
+      "loss": 0.7278,
+      "step": 840
+    },
+    {
+      "epoch": 0.3272692270670902,
+      "grad_norm": 0.2795485258102417,
+      "learning_rate": 7.861209523329494e-06,
+      "loss": 0.7404,
+      "step": 850
+    },
+    {
+      "epoch": 0.3311194532678795,
+      "grad_norm": 0.28989169001579285,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.7226,
+      "step": 860
+    },
+    {
+      "epoch": 0.3349696794686688,
+      "grad_norm": 0.3044799566268921,
+      "learning_rate": 7.758052511879035e-06,
+      "loss": 0.733,
+      "step": 870
+    },
+    {
+      "epoch": 0.3388199056694581,
+      "grad_norm": 0.2709113359451294,
+      "learning_rate": 7.705826503098642e-06,
+      "loss": 0.7296,
+      "step": 880
+    },
+    {
+      "epoch": 0.34267013187024736,
+      "grad_norm": 0.2792675197124481,
+      "learning_rate": 7.653179634243979e-06,
+      "loss": 0.7306,
+      "step": 890
+    },
+    {
+      "epoch": 0.34652035807103665,
+      "grad_norm": 0.26563265919685364,
+      "learning_rate": 7.600120093927696e-06,
+      "loss": 0.7318,
+      "step": 900
+    },
+    {
+      "epoch": 0.350370584271826,
+      "grad_norm": 0.3063053786754608,
+      "learning_rate": 7.546656134948729e-06,
+      "loss": 0.7202,
+      "step": 910
+    },
+    {
+      "epoch": 0.3542208104726153,
+      "grad_norm": 0.29810911417007446,
+      "learning_rate": 7.492796073008661e-06,
+      "loss": 0.7217,
+      "step": 920
+    },
+    {
+      "epoch": 0.35807103667340456,
+      "grad_norm": 0.299349844455719,
+      "learning_rate": 7.438548285418326e-06,
+      "loss": 0.7084,
+      "step": 930
+    },
+    {
+      "epoch": 0.36192126287419385,
+      "grad_norm": 0.37874674797058105,
+      "learning_rate": 7.3839212097948e-06,
+      "loss": 0.743,
+      "step": 940
+    },
+    {
+      "epoch": 0.36577148907498314,
+      "grad_norm": 0.28563326597213745,
+      "learning_rate": 7.328923342749039e-06,
+      "loss": 0.7559,
+      "step": 950
+    },
+    {
+      "epoch": 0.3696217152757725,
+      "grad_norm": 0.2894061207771301,
+      "learning_rate": 7.273563238564319e-06,
+      "loss": 0.7421,
+      "step": 960
+    },
+    {
+      "epoch": 0.37347194147656176,
+      "grad_norm": 0.276836097240448,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.7236,
+      "step": 970
+    },
+    {
+      "epoch": 0.37732216767735105,
+      "grad_norm": 0.2859756350517273,
+      "learning_rate": 7.161790816280858e-06,
+      "loss": 0.7283,
+      "step": 980
+    },
+    {
+      "epoch": 0.38117239387814034,
+      "grad_norm": 0.2969167232513428,
+      "learning_rate": 7.1053958830920125e-06,
+      "loss": 0.7106,
+      "step": 990
+    },
+    {
+      "epoch": 0.3850226200789296,
+      "grad_norm": 0.2738693058490753,
+      "learning_rate": 7.0486734798799715e-06,
+      "loss": 0.7313,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3888728462797189,
+      "grad_norm": 0.2855803966522217,
+      "learning_rate": 6.991632429159708e-06,
+      "loss": 0.752,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39272307248050825,
+      "grad_norm": 0.30628353357315063,
+      "learning_rate": 6.934281603008132e-06,
+      "loss": 0.7431,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39657329868129754,
+      "grad_norm": 0.2854093313217163,
+      "learning_rate": 6.876629921684143e-06,
+      "loss": 0.7412,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4004235248820868,
+      "grad_norm": 0.2983599007129669,
+      "learning_rate": 6.818686352241199e-06,
+      "loss": 0.7334,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4042737510828761,
+      "grad_norm": 0.2753029465675354,
+      "learning_rate": 6.760459907132573e-06,
+      "loss": 0.7266,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4081239772836654,
+      "grad_norm": 0.2707870304584503,
+      "learning_rate": 6.701959642809589e-06,
+      "loss": 0.7316,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41197420348445474,
+      "grad_norm": 0.2813275158405304,
+      "learning_rate": 6.643194658312987e-06,
+      "loss": 0.7265,
+      "step": 1070
+    },
+    {
+      "epoch": 0.415824429685244,
+      "grad_norm": 0.2785429060459137,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.7487,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4196746558860333,
+      "grad_norm": 0.3292093873023987,
+      "learning_rate": 6.5249071294110735e-06,
+      "loss": 0.7325,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4235248820868226,
+      "grad_norm": 0.2846204936504364,
+      "learning_rate": 6.465402983265274e-06,
+      "loss": 0.7442,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4273751082876119,
+      "grad_norm": 0.27908071875572205,
+      "learning_rate": 6.405670910603243e-06,
+      "loss": 0.741,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43122533448840117,
+      "grad_norm": 0.318387508392334,
+      "learning_rate": 6.345720202059291e-06,
+      "loss": 0.737,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4350755606891905,
+      "grad_norm": 0.29245486855506897,
+      "learning_rate": 6.285560182274008e-06,
+      "loss": 0.717,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4389257868899798,
+      "grad_norm": 0.27780839800834656,
+      "learning_rate": 6.225200208443936e-06,
+      "loss": 0.7383,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4427760130907691,
+      "grad_norm": 0.26311397552490234,
+      "learning_rate": 6.16464966886616e-06,
+      "loss": 0.7265,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44662623929155837,
+      "grad_norm": 0.32639428973197937,
+      "learning_rate": 6.103917981478065e-06,
+      "loss": 0.7482,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45047646549234766,
+      "grad_norm": 0.28023308515548706,
+      "learning_rate": 6.043014592392489e-06,
+      "loss": 0.7619,
+      "step": 1170
+    },
+    {
+      "epoch": 0.454326691693137,
+      "grad_norm": 0.2675595283508301,
+      "learning_rate": 5.981948974428485e-06,
+      "loss": 0.7354,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4581769178939263,
+      "grad_norm": 0.2984084486961365,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.713,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46202714409471557,
+      "grad_norm": 0.315872460603714,
+      "learning_rate": 5.8593690678282355e-06,
+      "loss": 0.7291,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46587737029550486,
+      "grad_norm": 0.2873520255088806,
+      "learning_rate": 5.797873845081301e-06,
+      "loss": 0.7326,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46972759649629414,
+      "grad_norm": 0.3038581311702728,
+      "learning_rate": 5.736254522269075e-06,
+      "loss": 0.7465,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47357782269708343,
+      "grad_norm": 0.284284383058548,
+      "learning_rate": 5.6745206835658386e-06,
+      "loss": 0.7495,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47742804889787277,
+      "grad_norm": 0.2759275734424591,
+      "learning_rate": 5.612681930957495e-06,
+      "loss": 0.7207,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48127827509866206,
+      "grad_norm": 0.26812613010406494,
+      "learning_rate": 5.550747882748095e-06,
+      "loss": 0.718,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48512850129945134,
+      "grad_norm": 0.2943764626979828,
+      "learning_rate": 5.4887281720638195e-06,
+      "loss": 0.7248,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48897872750024063,
+      "grad_norm": 0.31783831119537354,
+      "learning_rate": 5.42663244535466e-06,
+      "loss": 0.7106,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4928289537010299,
+      "grad_norm": 0.2942078113555908,
+      "learning_rate": 5.3644703608940205e-06,
+      "loss": 0.7198,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49667917990181926,
+      "grad_norm": 0.28230786323547363,
+      "learning_rate": 5.302251587276491e-06,
+      "loss": 0.7288,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5005294061026085,
+      "grad_norm": 0.2830665409564972,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.7297,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5043796323033978,
+      "grad_norm": 0.28477171063423157,
+      "learning_rate": 5.177682689530615e-06,
+      "loss": 0.7433,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5082298585041871,
+      "grad_norm": 0.28444939851760864,
+      "learning_rate": 5.11535194065619e-06,
+      "loss": 0.7127,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5120800847049765,
+      "grad_norm": 0.3156488835811615,
+      "learning_rate": 5.053003250119113e-06,
+      "loss": 0.7207,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5159303109057657,
+      "grad_norm": 0.2817505896091461,
+      "learning_rate": 4.9906463155383925e-06,
+      "loss": 0.7282,
+      "step": 1340
+    },
+    {
+      "epoch": 0.519780537106555,
+      "grad_norm": 0.282015860080719,
+      "learning_rate": 4.9282908358153046e-06,
+      "loss": 0.7253,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5236307633073443,
+      "grad_norm": 0.30179426074028015,
+      "learning_rate": 4.8659465096248345e-06,
+      "loss": 0.7387,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5274809895081336,
+      "grad_norm": 0.31428253650665283,
+      "learning_rate": 4.803623033907169e-06,
+      "loss": 0.736,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5313312157089229,
+      "grad_norm": 0.27614450454711914,
+      "learning_rate": 4.741330102359441e-06,
+      "loss": 0.7427,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5351814419097122,
+      "grad_norm": 0.3133978545665741,
+      "learning_rate": 4.679077403927996e-06,
+      "loss": 0.7224,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5390316681105015,
+      "grad_norm": 0.28913190960884094,
+      "learning_rate": 4.616874621301373e-06,
+      "loss": 0.7256,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5428818943112907,
+      "grad_norm": 0.2915271818637848,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.7356,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5467321205120801,
+      "grad_norm": 0.31563112139701843,
+      "learning_rate": 4.492657493892822e-06,
+      "loss": 0.722,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5505823467128694,
+      "grad_norm": 0.3104857802391052,
+      "learning_rate": 4.430662469650999e-06,
+      "loss": 0.7361,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5544325729136587,
+      "grad_norm": 0.30692487955093384,
+      "learning_rate": 4.368755999289118e-06,
+      "loss": 0.7405,
+      "step": 1440
+    },
+    {
+      "epoch": 0.558282799114448,
+      "grad_norm": 0.3223426043987274,
+      "learning_rate": 4.306947711643949e-06,
+      "loss": 0.7252,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5621330253152372,
+      "grad_norm": 0.2885618507862091,
+      "learning_rate": 4.245247220281068e-06,
+      "loss": 0.7286,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5659832515160266,
+      "grad_norm": 0.2694733738899231,
+      "learning_rate": 4.183664121999587e-06,
+      "loss": 0.7254,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5698334777168159,
+      "grad_norm": 0.28825873136520386,
+      "learning_rate": 4.122207995339474e-06,
+      "loss": 0.7241,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5736837039176051,
+      "grad_norm": 0.27138063311576843,
+      "learning_rate": 4.060888399091734e-06,
+      "loss": 0.726,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5775339301183945,
+      "grad_norm": 0.27284517884254456,
+      "learning_rate": 3.999714870811639e-06,
+      "loss": 0.7357,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5813841563191837,
+      "grad_norm": 0.2929173409938812,
+      "learning_rate": 3.938696925335275e-06,
+      "loss": 0.7307,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5852343825199731,
+      "grad_norm": 0.27290216088294983,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.7369,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5890846087207623,
+      "grad_norm": 0.288068950176239,
+      "learning_rate": 3.817165719666382e-06,
+      "loss": 0.7387,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5929348349215516,
+      "grad_norm": 0.26924553513526917,
+      "learning_rate": 3.756671362249841e-06,
+      "loss": 0.7415,
+      "step": 1540
+    },
+    {
+      "epoch": 0.596785061122341,
+      "grad_norm": 0.27286291122436523,
+      "learning_rate": 3.6963703902488907e-06,
+      "loss": 0.7298,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6006352873231302,
+      "grad_norm": 0.2768002450466156,
+      "learning_rate": 3.6362721827835655e-06,
+      "loss": 0.7437,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6044855135239195,
+      "grad_norm": 0.30911582708358765,
+      "learning_rate": 3.576386087436212e-06,
+      "loss": 0.7408,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6083357397247088,
+      "grad_norm": 0.305303692817688,
+      "learning_rate": 3.5167214187975913e-06,
+      "loss": 0.7314,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6121859659254981,
+      "grad_norm": 0.24806715548038483,
+      "learning_rate": 3.457287457018086e-06,
+      "loss": 0.7332,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6160361921262875,
+      "grad_norm": 0.2628382444381714,
+      "learning_rate": 3.3980934463643003e-06,
+      "loss": 0.721,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6198864183270767,
+      "grad_norm": 0.26443472504615784,
+      "learning_rate": 3.3391485937812053e-06,
+      "loss": 0.7042,
+      "step": 1610
+    },
+    {
+      "epoch": 0.623736644527866,
+      "grad_norm": 0.29239121079444885,
+      "learning_rate": 3.280462067460115e-06,
+      "loss": 0.704,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6275868707286553,
+      "grad_norm": 0.277095764875412,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.7338,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6314370969294446,
+      "grad_norm": 0.2595994174480438,
+      "learning_rate": 3.163900464051089e-06,
+      "loss": 0.7262,
+      "step": 1640
+    },
+    {
+      "epoch": 0.635287323130234,
+      "grad_norm": 0.2865414023399353,
+      "learning_rate": 3.1060435167748826e-06,
+      "loss": 0.7348,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6391375493310232,
+      "grad_norm": 0.30455315113067627,
+      "learning_rate": 3.048481152564253e-06,
+      "loss": 0.7155,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6429877755318125,
+      "grad_norm": 0.2905813455581665,
+      "learning_rate": 2.9912223245804017e-06,
+      "loss": 0.7381,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6468380017326018,
+      "grad_norm": 0.2909297049045563,
+      "learning_rate": 2.934275938772981e-06,
+      "loss": 0.7294,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6506882279333911,
+      "grad_norm": 0.3035866916179657,
+      "learning_rate": 2.8776508524948676e-06,
+      "loss": 0.7225,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6545384541341804,
+      "grad_norm": 0.25847944617271423,
+      "learning_rate": 2.821355873124503e-06,
+      "loss": 0.7194,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6583886803349697,
+      "grad_norm": 0.28780606389045715,
+      "learning_rate": 2.7653997566960144e-06,
+      "loss": 0.7346,
+      "step": 1710
+    },
+    {
+      "epoch": 0.662238906535759,
+      "grad_norm": 0.2938503324985504,
+      "learning_rate": 2.709791206537309e-06,
+      "loss": 0.7244,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6660891327365482,
+      "grad_norm": 0.28053998947143555,
+      "learning_rate": 2.6545388719163747e-06,
+      "loss": 0.7267,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6699393589373376,
+      "grad_norm": 0.27124014496803284,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.7279,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6737895851381268,
+      "grad_norm": 0.2700261175632477,
+      "learning_rate": 2.545137167997007e-06,
+      "loss": 0.7229,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6776398113389162,
+      "grad_norm": 0.2735235095024109,
+      "learning_rate": 2.4910048148705972e-06,
+      "loss": 0.713,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6814900375397055,
+      "grad_norm": 0.30441612005233765,
+      "learning_rate": 2.4372627069793327e-06,
+      "loss": 0.7578,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6853402637404947,
+      "grad_norm": 0.2875123620033264,
+      "learning_rate": 2.3839192032876486e-06,
+      "loss": 0.7406,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6891904899412841,
+      "grad_norm": 0.27900081872940063,
+      "learning_rate": 2.3309826007617047e-06,
+      "loss": 0.7512,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6930407161420733,
+      "grad_norm": 0.2807946503162384,
+      "learning_rate": 2.278461133078881e-06,
+      "loss": 0.7188,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6968909423428626,
+      "grad_norm": 0.2975514829158783,
+      "learning_rate": 2.2263629693471258e-06,
+      "loss": 0.7085,
+      "step": 1810
+    },
+    {
+      "epoch": 0.700741168543652,
+      "grad_norm": 0.28560343384742737,
+      "learning_rate": 2.1746962128343414e-06,
+      "loss": 0.7309,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7045913947444412,
+      "grad_norm": 0.2825625240802765,
+      "learning_rate": 2.1234688997080183e-06,
+      "loss": 0.7202,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7084416209452306,
+      "grad_norm": 0.30131927132606506,
+      "learning_rate": 2.0726889977853053e-06,
+      "loss": 0.7249,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7122918471460198,
+      "grad_norm": 0.2992495894432068,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.7383,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7161420733468091,
+      "grad_norm": 0.2704795002937317,
+      "learning_rate": 1.9725029496425814e-06,
+      "loss": 0.7321,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7199922995475985,
+      "grad_norm": 0.2691941261291504,
+      "learning_rate": 1.923112386205727e-06,
+      "loss": 0.7265,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7238425257483877,
+      "grad_norm": 0.2685709297657013,
+      "learning_rate": 1.874200397115079e-06,
+      "loss": 0.7348,
+      "step": 1880
+    },
+    {
+      "epoch": 0.727692751949177,
+      "grad_norm": 0.28488531708717346,
+      "learning_rate": 1.825774590065857e-06,
+      "loss": 0.7208,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7315429781499663,
+      "grad_norm": 0.2895171344280243,
+      "learning_rate": 1.7778424971332813e-06,
+      "loss": 0.7269,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7353932043507556,
+      "grad_norm": 0.2778487503528595,
+      "learning_rate": 1.7304115736010407e-06,
+      "loss": 0.7258,
+      "step": 1910
+    },
+    {
+      "epoch": 0.739243430551545,
+      "grad_norm": 0.28875255584716797,
+      "learning_rate": 1.6834891968017108e-06,
+      "loss": 0.7199,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7430936567523342,
+      "grad_norm": 0.2968583106994629,
+      "learning_rate": 1.6370826649692884e-06,
+      "loss": 0.735,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7469438829531235,
+      "grad_norm": 0.28970542550086975,
+      "learning_rate": 1.5911991961040463e-06,
+      "loss": 0.743,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7507941091539128,
+      "grad_norm": 0.28740939497947693,
+      "learning_rate": 1.545845926849851e-06,
+      "loss": 0.7309,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7546443353547021,
+      "grad_norm": 0.29496464133262634,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.7457,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7584945615554913,
+      "grad_norm": 0.298119455575943,
+      "learning_rate": 1.4567581203207238e-06,
+      "loss": 0.7389,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7623447877562807,
+      "grad_norm": 0.2869229316711426,
+      "learning_rate": 1.413037439625593e-06,
+      "loss": 0.734,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76619501395707,
+      "grad_norm": 0.3105246424674988,
+      "learning_rate": 1.3698746695458847e-06,
+      "loss": 0.7294,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7700452401578592,
+      "grad_norm": 0.27054670453071594,
+      "learning_rate": 1.3272765235521745e-06,
+      "loss": 0.7218,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7738954663586486,
+      "grad_norm": 0.27724507451057434,
+      "learning_rate": 1.2852496272942877e-06,
+      "loss": 0.7313,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7777456925594378,
+      "grad_norm": 0.2845175564289093,
+      "learning_rate": 1.2438005175707396e-06,
+      "loss": 0.7189,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7815959187602272,
+      "grad_norm": 0.28450119495391846,
+      "learning_rate": 1.2029356413120297e-06,
+      "loss": 0.7046,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7854461449610165,
+      "grad_norm": 0.2713913023471832,
+      "learning_rate": 1.1626613545778786e-06,
+      "loss": 0.7206,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7892963711618057,
+      "grad_norm": 0.2675632834434509,
+      "learning_rate": 1.1229839215686294e-06,
+      "loss": 0.7249,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7931465973625951,
+      "grad_norm": 0.2641056478023529,
+      "learning_rate": 1.0839095136509188e-06,
+      "loss": 0.7092,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7969968235633843,
+      "grad_norm": 0.30134788155555725,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.7287,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8008470497641736,
+      "grad_norm": 0.28334107995033264,
+      "learning_rate": 1.007593988643402e-06,
+      "loss": 0.7137,
+      "step": 2080
+    },
+    {
+      "epoch": 0.804697275964963,
+      "grad_norm": 0.29915478825569153,
+      "learning_rate": 9.703647415524613e-07,
+      "loss": 0.7332,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8085475021657522,
+      "grad_norm": 0.2817022204399109,
+      "learning_rate": 9.337622577045485e-07,
+      "loss": 0.7231,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8123977283665416,
+      "grad_norm": 0.2891615331172943,
+      "learning_rate": 8.977922301934588e-07,
+      "loss": 0.7324,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8162479545673308,
+      "grad_norm": 0.2778261601924896,
+      "learning_rate": 8.624602537417009e-07,
+      "loss": 0.7342,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8200981807681201,
+      "grad_norm": 0.30798113346099854,
+      "learning_rate": 8.277718238303123e-07,
+      "loss": 0.7222,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8239484069689095,
+      "grad_norm": 0.27277201414108276,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.7232,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8277986331696987,
+      "grad_norm": 0.300516277551651,
+      "learning_rate": 7.603470842324223e-07,
+      "loss": 0.7161,
+      "step": 2150
+    },
+    {
+      "epoch": 0.831648859370488,
+      "grad_norm": 0.2740214169025421,
+      "learning_rate": 7.276212616857575e-07,
+      "loss": 0.7349,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8354990855712773,
+      "grad_norm": 0.27276477217674255,
+      "learning_rate": 6.955599583279837e-07,
+      "loss": 0.7309,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8393493117720666,
+      "grad_norm": 0.2759648859500885,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.7253,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8431995379728559,
+      "grad_norm": 0.2740418016910553,
+      "learning_rate": 6.334507521075711e-07,
+      "loss": 0.7335,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8470497641736452,
+      "grad_norm": 0.2897599935531616,
+      "learning_rate": 6.034125096149179e-07,
+      "loss": 0.7279,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8508999903744345,
+      "grad_norm": 0.29967185854911804,
+      "learning_rate": 5.740581055485722e-07,
+      "loss": 0.7,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8547502165752238,
+      "grad_norm": 0.2607537806034088,
+      "learning_rate": 5.453921056471961e-07,
+      "loss": 0.7322,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8586004427760131,
+      "grad_norm": 0.2886735498905182,
+      "learning_rate": 5.174189685761288e-07,
+      "loss": 0.7254,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8624506689768023,
+      "grad_norm": 0.2727143168449402,
+      "learning_rate": 4.901430452338979e-07,
+      "loss": 0.7364,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8663008951775917,
+      "grad_norm": 0.27437812089920044,
+      "learning_rate": 4.6356857807548286e-07,
+      "loss": 0.7217,
+      "step": 2250
+    },
+    {
+      "epoch": 0.870151121378381,
+      "grad_norm": 0.2662178874015808,
+      "learning_rate": 4.376997004524541e-07,
+      "loss": 0.7111,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8740013475791703,
+      "grad_norm": 0.2852674424648285,
+      "learning_rate": 4.12540435970073e-07,
+      "loss": 0.7196,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8778515737799596,
+      "grad_norm": 0.2781769335269928,
+      "learning_rate": 3.880946978614703e-07,
+      "loss": 0.7319,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8817017999807488,
+      "grad_norm": 0.2767511010169983,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.732,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8855520261815382,
+      "grad_norm": 0.2681233584880829,
+      "learning_rate": 3.41358898202781e-07,
+      "loss": 0.713,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8894022523823275,
+      "grad_norm": 0.27551859617233276,
+      "learning_rate": 3.190761058667741e-07,
+      "loss": 0.7228,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8932524785831167,
+      "grad_norm": 0.2847493886947632,
+      "learning_rate": 2.9752137720206463e-07,
+      "loss": 0.7285,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8971027047839061,
+      "grad_norm": 0.2679632604122162,
+      "learning_rate": 2.7669806479785067e-07,
+      "loss": 0.7341,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9009529309846953,
+      "grad_norm": 0.2965357005596161,
+      "learning_rate": 2.566094074799763e-07,
+      "loss": 0.7262,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9048031571854847,
+      "grad_norm": 0.29054203629493713,
+      "learning_rate": 2.3725852980716557e-07,
+      "loss": 0.7123,
+      "step": 2350
+    },
+    {
+      "epoch": 0.908653383386274,
+      "grad_norm": 0.26517704129219055,
+      "learning_rate": 2.1864844158503983e-07,
+      "loss": 0.7225,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9125036095870632,
+      "grad_norm": 0.3085063397884369,
+      "learning_rate": 2.007820373979702e-07,
+      "loss": 0.7159,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9163538357878526,
+      "grad_norm": 0.2791442275047302,
+      "learning_rate": 1.8366209615886532e-07,
+      "loss": 0.7243,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9202040619886418,
+      "grad_norm": 0.2880585193634033,
+      "learning_rate": 1.6729128067693613e-07,
+      "loss": 0.7364,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9240542881894311,
+      "grad_norm": 0.3002503514289856,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.7351,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9279045143902205,
+      "grad_norm": 0.2781367003917694,
+      "learning_rate": 1.3680709523610058e-07,
+      "loss": 0.7202,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9317547405910097,
+      "grad_norm": 0.2679388225078583,
+      "learning_rate": 1.2269846674030418e-07,
+      "loss": 0.7393,
+      "step": 2420
+    },
+    {
+      "epoch": 0.935604966791799,
+      "grad_norm": 0.2897999882698059,
+      "learning_rate": 1.0934844619042595e-07,
+      "loss": 0.738,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9394551929925883,
+      "grad_norm": 0.29413822293281555,
+      "learning_rate": 9.675911002803961e-08,
+      "loss": 0.7251,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9433054191933776,
+      "grad_norm": 0.3167572021484375,
+      "learning_rate": 8.493241637904381e-08,
+      "loss": 0.722,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9471556453941669,
+      "grad_norm": 0.2808409035205841,
+      "learning_rate": 7.387020474909689e-08,
+      "loss": 0.742,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9510058715949562,
+      "grad_norm": 0.2925013303756714,
+      "learning_rate": 6.357419573750856e-08,
+      "loss": 0.7356,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9548560977957455,
+      "grad_norm": 0.29145482182502747,
+      "learning_rate": 5.4045990769613365e-08,
+      "loss": 0.7194,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9587063239965348,
+      "grad_norm": 0.28070008754730225,
+      "learning_rate": 4.5287071847694434e-08,
+      "loss": 0.7258,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9625565501973241,
+      "grad_norm": 0.2796279788017273,
+      "learning_rate": 3.7298801320467817e-08,
+      "loss": 0.7205,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9664067763981133,
+      "grad_norm": 0.2745845317840576,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.7146,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9702570025989027,
+      "grad_norm": 0.29075002670288086,
+      "learning_rate": 2.3639055324410355e-08,
+      "loss": 0.7378,
+      "step": 2520
+    },
+    {
+      "epoch": 0.974107228799692,
+      "grad_norm": 0.2809404134750366,
+      "learning_rate": 1.7969704471369477e-08,
+      "loss": 0.7193,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9779574550004813,
+      "grad_norm": 0.2677890360355377,
+      "learning_rate": 1.3075250914142413e-08,
+      "loss": 0.7213,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9818076812012706,
+      "grad_norm": 0.26801350712776184,
+      "learning_rate": 8.956455928476049e-09,
+      "loss": 0.7298,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9856579074020598,
+      "grad_norm": 0.2880041003227234,
+      "learning_rate": 5.6139601453786275e-09,
+      "loss": 0.725,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9895081336028492,
+      "grad_norm": 0.2762836515903473,
+      "learning_rate": 3.0482834514805602e-09,
+      "loss": 0.7131,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9933583598036385,
+      "grad_norm": 0.26511457562446594,
+      "learning_rate": 1.2598249081713366e-09,
+      "loss": 0.7065,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9972085860044277,
+      "grad_norm": 0.2868124842643738,
+      "learning_rate": 2.488626895302826e-10,
+      "loss": 0.7382,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9999037443449803,
+      "step": 2597,
+      "total_flos": 4.3217624287286395e+18,
+      "train_loss": 0.8101910830553927,
+      "train_runtime": 12998.8779,
+      "train_samples_per_second": 51.151,
+      "train_steps_per_second": 0.2
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2597,
+  "num_input_tokens_seen": 418855352,
+  "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": 4.3217624287286395e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3base/mc2_l14_224/finetune/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c54be4cc87d9a57a009158c20630da657374b05a98e122ae0231707538fb0263
+size 6264

continuous/qwen3base/mc2_l14_224/pretrain/backup/config.json

@@ -0,0 +1,89 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1024,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "flat",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_tunable_parts": "mm_mlp_adapter",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "open_clip_hub:ViT-L-14",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "pad_token_id": 151643,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "transformers_version": "4.57.6",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": "/cache/ckpt/download/tokenizer/continuous/mc2/mc2_l14_224.pt",
+  "vocab_size": 151936
+}

continuous/qwen3base/mc2_l14_224/pretrain/backup/mm_projector.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bdb1235ab11dc465cafbecfddd28bb5193cdf70dda3fe07269ec5adba0bb7c3d
+size 25184312

continuous/qwen3base/mc2_l14_224/pretrain/backup/trainer_state.json

@@ -0,0 +1,421 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 545,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01834862385321101,
+      "grad_norm": 2.539327383041382,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 7.0099,
+      "step": 10
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 0.3616117537021637,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 5.469,
+      "step": 20
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.21763159334659576,
+      "learning_rate": 0.0009999907688176173,
+      "loss": 5.2228,
+      "step": 30
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.10757898539304733,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 5.0901,
+      "step": 40
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.19080746173858643,
+      "learning_rate": 0.0009959345572940772,
+      "loss": 4.9962,
+      "step": 50
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.1693064570426941,
+      "learning_rate": 0.0009911550081155983,
+      "loss": 4.92,
+      "step": 60
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.19689656794071198,
+      "learning_rate": 0.0009845624347566433,
+      "loss": 4.8439,
+      "step": 70
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.16576434671878815,
+      "learning_rate": 0.0009761811727013547,
+      "loss": 4.7914,
+      "step": 80
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.13808757066726685,
+      "learning_rate": 0.0009660421601067666,
+      "loss": 4.7262,
+      "step": 90
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.14857473969459534,
+      "learning_rate": 0.0009541828235992926,
+      "loss": 4.6819,
+      "step": 100
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.10433395206928253,
+      "learning_rate": 0.0009406469401200151,
+      "loss": 4.64,
+      "step": 110
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.14295491576194763,
+      "learning_rate": 0.0009254844753287492,
+      "loss": 4.6165,
+      "step": 120
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.178766131401062,
+      "learning_rate": 0.0009087513991633924,
+      "loss": 4.5792,
+      "step": 130
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.14406315982341766,
+      "learning_rate": 0.0008905094792353917,
+      "loss": 4.5526,
+      "step": 140
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.14694775640964508,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 4.5298,
+      "step": 150
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.13050810992717743,
+      "learning_rate": 0.0008497737783108237,
+      "loss": 4.4836,
+      "step": 160
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.10487756878137589,
+      "learning_rate": 0.0008274303669726426,
+      "loss": 4.4374,
+      "step": 170
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.18346159160137177,
+      "learning_rate": 0.0008038782961218137,
+      "loss": 4.4099,
+      "step": 180
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.13711173832416534,
+      "learning_rate": 0.000779204504653894,
+      "loss": 4.3872,
+      "step": 190
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.12244565039873123,
+      "learning_rate": 0.0007535000721258791,
+      "loss": 4.3567,
+      "step": 200
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.13075290620326996,
+      "learning_rate": 0.00072685988254984,
+      "loss": 4.3454,
+      "step": 210
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.10285156220197678,
+      "learning_rate": 0.0006993822741429906,
+      "loss": 4.3533,
+      "step": 220
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.10759589821100235,
+      "learning_rate": 0.0006711686763270818,
+      "loss": 4.3394,
+      "step": 230
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.11504572629928589,
+      "learning_rate": 0.0006423232353170831,
+      "loss": 4.3251,
+      "step": 240
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.10616423934698105,
+      "learning_rate": 0.0006129524296812335,
+      "loss": 4.3174,
+      "step": 250
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.10588352382183075,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 4.3208,
+      "step": 260
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.14252537488937378,
+      "learning_rate": 0.0005530699351157809,
+      "loss": 4.3006,
+      "step": 270
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.12065362185239792,
+      "learning_rate": 0.0005227792933277943,
+      "loss": 4.2957,
+      "step": 280
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.10242806375026703,
+      "learning_rate": 0.0004924045652352092,
+      "loss": 4.294,
+      "step": 290
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.10092831403017044,
+      "learning_rate": 0.0004620578745374633,
+      "loss": 4.2884,
+      "step": 300
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.09821973741054535,
+      "learning_rate": 0.0004318512414382059,
+      "loss": 4.3004,
+      "step": 310
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.10068673640489578,
+      "learning_rate": 0.0004018961691397148,
+      "loss": 4.2868,
+      "step": 320
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.09494099766016006,
+      "learning_rate": 0.0003723032322457458,
+      "loss": 4.2769,
+      "step": 330
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.1599879115819931,
+      "learning_rate": 0.0003431816685921625,
+      "loss": 4.2739,
+      "step": 340
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.1127098798751831,
+      "learning_rate": 0.0003146389760120416,
+      "loss": 4.275,
+      "step": 350
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.12905637919902802,
+      "learning_rate": 0.00028678051552373485,
+      "loss": 4.2826,
+      "step": 360
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.10942653566598892,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 4.2678,
+      "step": 370
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.10678768903017044,
+      "learning_rate": 0.0002335247266004629,
+      "loss": 4.279,
+      "step": 380
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.10701417922973633,
+      "learning_rate": 0.000208323983828828,
+      "loss": 4.2657,
+      "step": 390
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.10613582283258438,
+      "learning_rate": 0.0001841999188095224,
+      "loss": 4.2637,
+      "step": 400
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.10923099517822266,
+      "learning_rate": 0.00016124158186777676,
+      "loss": 4.2658,
+      "step": 410
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.09235149621963501,
+      "learning_rate": 0.0001395337202205211,
+      "loss": 4.258,
+      "step": 420
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.09309151023626328,
+      "learning_rate": 0.00011915646514489292,
+      "loss": 4.2622,
+      "step": 430
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.09392004460096359,
+      "learning_rate": 0.00010018503618578818,
+      "loss": 4.2673,
+      "step": 440
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.09320205450057983,
+      "learning_rate": 8.268946349432583e-05,
+      "loss": 4.2563,
+      "step": 450
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.09810817986726761,
+      "learning_rate": 6.673432932217117e-05,
+      "loss": 4.2545,
+      "step": 460
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.09133413434028625,
+      "learning_rate": 5.2378529625954684e-05,
+      "loss": 4.2388,
+      "step": 470
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.09453420341014862,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 4.2456,
+      "step": 480
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.08878912776708603,
+      "learning_rate": 2.8670803372379538e-05,
+      "loss": 4.2466,
+      "step": 490
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.0911199301481247,
+      "learning_rate": 1.9406390288876586e-05,
+      "loss": 4.2597,
+      "step": 500
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.09172216057777405,
+      "learning_rate": 1.191601558631461e-05,
+      "loss": 4.2553,
+      "step": 510
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.08866582065820694,
+      "learning_rate": 6.2273288462559305e-06,
+      "loss": 4.2478,
+      "step": 520
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.09136947244405746,
+      "learning_rate": 2.361328992552314e-06,
+      "loss": 4.25,
+      "step": 530
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.0899653360247612,
+      "learning_rate": 3.322867770044202e-07,
+      "loss": 4.2479,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "step": 545,
+      "total_flos": 3.4196958409549414e+17,
+      "train_loss": 4.484416531204084,
+      "train_runtime": 5475.8799,
+      "train_samples_per_second": 101.925,
+      "train_steps_per_second": 0.1
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 545,
+  "num_input_tokens_seen": 33142912,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.4196958409549414e+17,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3base/mc2_l14_224/pretrain/config.json

@@ -0,0 +1,89 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1024,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "flat",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_tunable_parts": "mm_mlp_adapter",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "open_clip_hub:ViT-L-14",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "pad_token_id": 151643,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "transformers_version": "4.57.6",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": "/cache/ckpt/download/tokenizer/continuous/mc2/mc2_l14_224.pt",
+  "vocab_size": 151936
+}

continuous/qwen3base/mc2_l14_224/pretrain/mm_projector.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36bd266da293bdf04df511b9a35070062357da9ffd15f0c2388edf26f6d67aac
+size 25184312

continuous/qwen3base/mc2_l14_224/pretrain/trainer_state.json

@@ -0,0 +1,421 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 545,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01834862385321101,
+      "grad_norm": 2.4062647819519043,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 7.0084,
+      "step": 10
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 0.5195721983909607,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 5.4762,
+      "step": 20
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.23257683217525482,
+      "learning_rate": 0.0009999907688176173,
+      "loss": 5.2274,
+      "step": 30
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.10695258527994156,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 5.0911,
+      "step": 40
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.19655221700668335,
+      "learning_rate": 0.0009959345572940772,
+      "loss": 4.997,
+      "step": 50
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.10130443423986435,
+      "learning_rate": 0.0009911550081155983,
+      "loss": 4.919,
+      "step": 60
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.12917162477970123,
+      "learning_rate": 0.0009845624347566433,
+      "loss": 4.841,
+      "step": 70
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.16836223006248474,
+      "learning_rate": 0.0009761811727013547,
+      "loss": 4.7877,
+      "step": 80
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.10781051218509674,
+      "learning_rate": 0.0009660421601067666,
+      "loss": 4.7224,
+      "step": 90
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.12383826076984406,
+      "learning_rate": 0.0009541828235992926,
+      "loss": 4.676,
+      "step": 100
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.12736281752586365,
+      "learning_rate": 0.0009406469401200151,
+      "loss": 4.6336,
+      "step": 110
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.15954840183258057,
+      "learning_rate": 0.0009254844753287492,
+      "loss": 4.6101,
+      "step": 120
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.12569516897201538,
+      "learning_rate": 0.0009087513991633924,
+      "loss": 4.5753,
+      "step": 130
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.11372912675142288,
+      "learning_rate": 0.0008905094792353917,
+      "loss": 4.5418,
+      "step": 140
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.1534566432237625,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 4.5222,
+      "step": 150
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.13649269938468933,
+      "learning_rate": 0.0008497737783108237,
+      "loss": 4.4709,
+      "step": 160
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.10703279823064804,
+      "learning_rate": 0.0008274303669726426,
+      "loss": 4.4267,
+      "step": 170
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.1435266137123108,
+      "learning_rate": 0.0008038782961218137,
+      "loss": 4.4014,
+      "step": 180
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.13080883026123047,
+      "learning_rate": 0.000779204504653894,
+      "loss": 4.3808,
+      "step": 190
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.10688038170337677,
+      "learning_rate": 0.0007535000721258791,
+      "loss": 4.3516,
+      "step": 200
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.11435964703559875,
+      "learning_rate": 0.00072685988254984,
+      "loss": 4.3424,
+      "step": 210
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.11136471480131149,
+      "learning_rate": 0.0006993822741429906,
+      "loss": 4.3512,
+      "step": 220
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.11484213173389435,
+      "learning_rate": 0.0006711686763270818,
+      "loss": 4.3384,
+      "step": 230
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.14153814315795898,
+      "learning_rate": 0.0006423232353170831,
+      "loss": 4.3241,
+      "step": 240
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.10687582194805145,
+      "learning_rate": 0.0006129524296812335,
+      "loss": 4.3155,
+      "step": 250
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.10379134863615036,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 4.3196,
+      "step": 260
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.14331601560115814,
+      "learning_rate": 0.0005530699351157809,
+      "loss": 4.3005,
+      "step": 270
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.11909300088882446,
+      "learning_rate": 0.0005227792933277943,
+      "loss": 4.2943,
+      "step": 280
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.1136595755815506,
+      "learning_rate": 0.0004924045652352092,
+      "loss": 4.2932,
+      "step": 290
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.11162654310464859,
+      "learning_rate": 0.0004620578745374633,
+      "loss": 4.2896,
+      "step": 300
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.11288333684206009,
+      "learning_rate": 0.0004318512414382059,
+      "loss": 4.3,
+      "step": 310
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.10413885116577148,
+      "learning_rate": 0.0004018961691397148,
+      "loss": 4.2874,
+      "step": 320
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.1007218211889267,
+      "learning_rate": 0.0003723032322457458,
+      "loss": 4.2756,
+      "step": 330
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.15261238813400269,
+      "learning_rate": 0.0003431816685921625,
+      "loss": 4.2737,
+      "step": 340
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.12353686988353729,
+      "learning_rate": 0.0003146389760120416,
+      "loss": 4.2744,
+      "step": 350
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.14288808405399323,
+      "learning_rate": 0.00028678051552373485,
+      "loss": 4.2817,
+      "step": 360
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.11191391199827194,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 4.2678,
+      "step": 370
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.10162171721458435,
+      "learning_rate": 0.0002335247266004629,
+      "loss": 4.2785,
+      "step": 380
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.10101993381977081,
+      "learning_rate": 0.000208323983828828,
+      "loss": 4.2655,
+      "step": 390
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.1061282604932785,
+      "learning_rate": 0.0001841999188095224,
+      "loss": 4.263,
+      "step": 400
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.10576585680246353,
+      "learning_rate": 0.00016124158186777676,
+      "loss": 4.2666,
+      "step": 410
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.09205016493797302,
+      "learning_rate": 0.0001395337202205211,
+      "loss": 4.2587,
+      "step": 420
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.09878375381231308,
+      "learning_rate": 0.00011915646514489292,
+      "loss": 4.2626,
+      "step": 430
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.09508088231086731,
+      "learning_rate": 0.00010018503618578818,
+      "loss": 4.2686,
+      "step": 440
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.09292156994342804,
+      "learning_rate": 8.268946349432583e-05,
+      "loss": 4.2557,
+      "step": 450
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.09302239120006561,
+      "learning_rate": 6.673432932217117e-05,
+      "loss": 4.2549,
+      "step": 460
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.0887775868177414,
+      "learning_rate": 5.2378529625954684e-05,
+      "loss": 4.2379,
+      "step": 470
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.09808021038770676,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 4.2455,
+      "step": 480
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.08877656608819962,
+      "learning_rate": 2.8670803372379538e-05,
+      "loss": 4.2461,
+      "step": 490
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.09088163822889328,
+      "learning_rate": 1.9406390288876586e-05,
+      "loss": 4.2603,
+      "step": 500
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.09198262542486191,
+      "learning_rate": 1.191601558631461e-05,
+      "loss": 4.2553,
+      "step": 510
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.09195968508720398,
+      "learning_rate": 6.2273288462559305e-06,
+      "loss": 4.2475,
+      "step": 520
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.08898094296455383,
+      "learning_rate": 2.361328992552314e-06,
+      "loss": 4.2499,
+      "step": 530
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.09133780747652054,
+      "learning_rate": 3.322867770044202e-07,
+      "loss": 4.2489,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "step": 545,
+      "total_flos": 3.4196958409549414e+17,
+      "train_loss": 4.4826888303144266,
+      "train_runtime": 5393.7191,
+      "train_samples_per_second": 103.477,
+      "train_steps_per_second": 0.101
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 545,
+  "num_input_tokens_seen": 33142912,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.4196958409549414e+17,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3base/siglip2_g16_384/finetune/added_tokens.json

@@ -0,0 +1,28 @@
+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 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
+}

continuous/qwen3base/siglip2_g16_384/finetune/chat_template.jinja

@@ -0,0 +1,85 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\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 <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set content = message.content %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in message.content %}
+                {%- set content = message.content.split('</think>')[-1].lstrip('\n') %}
+                {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

continuous/qwen3base/siglip2_g16_384/finetune/config.json

@@ -0,0 +1,89 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1536,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "flat",
+  "mm_projector_lr": 0.0001,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_tunable_parts": "mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "open_clip_hub:ViT-gopt-16-SigLIP2-384",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "pad_token_id": 151643,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "transformers_version": "4.57.6",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": "/cache/ckpt/download/tokenizer/continuous/siglip2/siglip2_g16_384.npz",
+  "vocab_size": 151936
+}

continuous/qwen3base/siglip2_g16_384/finetune/generation_config.json

@@ -0,0 +1,8 @@
+{
+  "eos_token_id": [
+    151643
+  ],
+  "max_new_tokens": 2048,
+  "pad_token_id": 151643,
+  "transformers_version": "4.57.6"
+}

continuous/qwen3base/siglip2_g16_384/finetune/model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:264e01a27d2977800c690e0458dd84057f84e1cae0584bfbfa323282d6709671
+size 4994164072

continuous/qwen3base/siglip2_g16_384/finetune/model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87fdf0ce40c6d2caca22ed5582eb65ce8aab12f3a4c65ab65b51162efe828330
+size 1426115616

continuous/qwen3base/siglip2_g16_384/finetune/model.safetensors.index.json

@@ -0,0 +1,821 @@
+{
+  "metadata": {
+    "total_parameters": 2891578880,
+    "total_size": 6420175872
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00002-of-00002.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.mm_projector.0.bias": "model-00002-of-00002.safetensors",
+    "model.mm_projector.0.weight": "model-00002-of-00002.safetensors",
+    "model.mm_projector.2.bias": "model-00002-of-00002.safetensors",
+    "model.mm_projector.2.weight": "model-00002-of-00002.safetensors",
+    "model.norm.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.attn_pool.kv.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.attn_pool.kv.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.attn_pool.latent": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.attn_pool.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.attn_pool.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.attn_pool.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.attn_pool.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.attn_pool.norm.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.attn_pool.norm.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.attn_pool.proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.attn_pool.proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.attn_pool.q.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.attn_pool.q.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.0.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.0.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.0.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.0.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.0.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.0.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.0.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.0.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.0.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.0.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.0.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.0.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.1.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.1.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.1.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.1.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.1.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.1.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.1.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.1.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.1.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.1.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.1.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.1.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.10.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.10.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.10.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.10.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.10.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.10.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.10.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.10.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.10.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.10.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.10.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.10.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.11.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.11.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.11.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.11.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.11.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.11.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.11.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.11.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.11.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.11.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.11.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.11.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.12.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.12.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.12.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.12.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.12.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.12.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.12.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.12.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.12.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.12.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.12.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.12.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.13.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.13.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.13.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.13.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.13.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.13.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.13.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.13.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.13.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.13.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.13.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.13.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.14.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.14.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.14.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.14.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.14.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.14.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.14.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.14.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.14.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.14.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.14.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.14.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.15.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.15.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.15.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.15.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.15.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.15.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.15.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.15.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.15.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.15.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.15.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.15.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.16.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.16.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.16.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.16.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.16.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.16.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.16.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.16.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.16.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.16.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.16.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.16.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.17.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.17.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.17.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.17.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.17.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.17.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.17.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.17.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.17.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.17.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.17.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.17.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.18.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.18.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.18.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.18.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.18.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.18.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.18.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.18.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.18.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.18.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.18.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.18.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.19.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.19.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.19.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.19.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.19.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.19.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.19.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.19.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.19.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.19.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.19.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.19.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.2.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.2.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.2.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.2.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.2.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.2.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.2.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.2.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.2.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.2.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.2.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.2.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.20.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.20.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.20.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.20.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.20.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.20.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.20.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.20.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.20.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.20.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.20.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.20.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.21.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.21.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.21.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.21.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.21.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.21.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.21.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.21.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.21.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.21.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.21.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.21.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.22.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.22.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.22.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.22.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.22.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.22.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.22.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.22.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.22.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.22.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.22.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.22.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.23.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.23.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.23.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.23.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.23.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.23.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.23.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.23.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.23.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.23.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.23.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.23.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.24.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.24.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.24.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.24.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.24.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.24.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.24.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.24.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.24.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.24.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.24.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.24.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.25.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.25.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.25.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.25.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.25.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.25.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.25.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.25.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.25.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.25.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.25.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.25.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.26.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.26.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.26.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.26.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.26.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.26.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.26.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.26.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.26.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.26.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.26.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.26.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.27.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.27.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.27.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.27.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.27.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.27.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.27.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.27.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.27.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.27.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.27.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.27.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.28.attn.proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.28.attn.proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.28.attn.qkv.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.28.attn.qkv.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.28.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.28.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.28.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.28.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.28.norm1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.28.norm1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.28.norm2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.28.norm2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.29.attn.proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.29.attn.proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.29.attn.qkv.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.29.attn.qkv.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.29.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.29.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.29.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.29.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.29.norm1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.29.norm1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.29.norm2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.29.norm2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.3.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.3.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.3.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.3.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.3.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.3.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.3.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.3.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.3.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.3.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.3.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.3.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.30.attn.proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.30.attn.proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.30.attn.qkv.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.30.attn.qkv.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.30.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.30.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.30.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.30.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.30.norm1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.30.norm1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.30.norm2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.30.norm2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.31.attn.proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.31.attn.proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.31.attn.qkv.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.31.attn.qkv.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.31.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.31.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.31.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.31.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.31.norm1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.31.norm1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.31.norm2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.31.norm2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.32.attn.proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.32.attn.proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.32.attn.qkv.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.32.attn.qkv.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.32.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.32.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.32.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.32.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.32.norm1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.32.norm1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.32.norm2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.32.norm2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.33.attn.proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.33.attn.proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.33.attn.qkv.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.33.attn.qkv.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.33.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.33.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.33.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.33.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.33.norm1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.33.norm1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.33.norm2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.33.norm2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.34.attn.proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.34.attn.proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.34.attn.qkv.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.34.attn.qkv.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.34.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.34.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.34.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.34.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.34.norm1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.34.norm1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.34.norm2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.34.norm2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.35.attn.proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.35.attn.proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.35.attn.qkv.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.35.attn.qkv.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.35.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.35.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.35.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.35.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.35.norm1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.35.norm1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.35.norm2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.35.norm2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.36.attn.proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.36.attn.proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.36.attn.qkv.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.36.attn.qkv.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.36.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.36.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.36.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.36.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.36.norm1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.36.norm1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.36.norm2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.36.norm2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.37.attn.proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.37.attn.proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.37.attn.qkv.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.37.attn.qkv.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.37.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.37.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.37.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.37.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.37.norm1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.37.norm1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.37.norm2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.37.norm2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.38.attn.proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.38.attn.proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.38.attn.qkv.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.38.attn.qkv.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.38.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.38.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.38.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.38.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.38.norm1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.38.norm1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.38.norm2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.38.norm2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.39.attn.proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.39.attn.proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.39.attn.qkv.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.39.attn.qkv.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.39.mlp.fc1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.39.mlp.fc1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.39.mlp.fc2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.39.mlp.fc2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.39.norm1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.39.norm1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.39.norm2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.39.norm2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.4.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.4.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.4.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.4.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.4.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.4.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.4.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.4.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.4.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.4.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.4.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.4.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.5.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.5.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.5.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.5.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.5.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.5.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.5.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.5.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.5.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.5.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.5.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.5.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.6.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.6.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.6.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.6.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.6.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.6.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.6.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.6.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.6.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.6.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.6.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.6.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.7.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.7.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.7.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.7.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.7.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.7.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.7.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.7.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.7.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.7.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.7.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.7.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.8.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.8.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.8.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.8.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.8.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.8.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.8.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.8.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.8.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.8.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.8.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.8.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.9.attn.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.9.attn.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.9.attn.qkv.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.9.attn.qkv.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.9.mlp.fc1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.9.mlp.fc1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.9.mlp.fc2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.9.mlp.fc2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.9.norm1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.9.norm1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.9.norm2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.blocks.9.norm2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.norm.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.norm.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.patch_embed.proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.patch_embed.proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.trunk.pos_embed": "model-00001-of-00002.safetensors"
+  }
+}

continuous/qwen3base/siglip2_g16_384/finetune/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": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

continuous/qwen3base/siglip2_g16_384/finetune/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
+size 11422654

continuous/qwen3base/siglip2_g16_384/finetune/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": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "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": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "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": "<|endoftext|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

continuous/qwen3base/siglip2_g16_384/finetune/trainer_state.json

@@ -0,0 +1,1856 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999037443449803,
+  "eval_steps": 500,
+  "global_step": 2597,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003850226200789296,
+      "grad_norm": 3.2963039875030518,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 3.0682,
+      "step": 10
+    },
+    {
+      "epoch": 0.007700452401578592,
+      "grad_norm": 2.275623321533203,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 2.7701,
+      "step": 20
+    },
+    {
+      "epoch": 0.01155067860236789,
+      "grad_norm": 1.8229893445968628,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 2.4675,
+      "step": 30
+    },
+    {
+      "epoch": 0.015400904803157185,
+      "grad_norm": 1.612119436264038,
+      "learning_rate": 5e-06,
+      "loss": 2.3778,
+      "step": 40
+    },
+    {
+      "epoch": 0.019251131003946483,
+      "grad_norm": 1.4809908866882324,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 2.1928,
+      "step": 50
+    },
+    {
+      "epoch": 0.02310135720473578,
+      "grad_norm": 1.7272601127624512,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 2.1032,
+      "step": 60
+    },
+    {
+      "epoch": 0.026951583405525073,
+      "grad_norm": 1.7565045356750488,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 2.0093,
+      "step": 70
+    },
+    {
+      "epoch": 0.03080180960631437,
+      "grad_norm": 2.006140947341919,
+      "learning_rate": 9.999996111488723e-06,
+      "loss": 1.9438,
+      "step": 80
+    },
+    {
+      "epoch": 0.034652035807103666,
+      "grad_norm": 1.793285846710205,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 1.8106,
+      "step": 90
+    },
+    {
+      "epoch": 0.038502262007892966,
+      "grad_norm": 1.871764898300171,
+      "learning_rate": 9.998285264324193e-06,
+      "loss": 1.6886,
+      "step": 100
+    },
+    {
+      "epoch": 0.04235248820868226,
+      "grad_norm": 1.912412166595459,
+      "learning_rate": 9.996263605626052e-06,
+      "loss": 1.5689,
+      "step": 110
+    },
+    {
+      "epoch": 0.04620271440947156,
+      "grad_norm": 1.8161474466323853,
+      "learning_rate": 9.99346483580503e-06,
+      "loss": 1.4595,
+      "step": 120
+    },
+    {
+      "epoch": 0.05005294061026085,
+      "grad_norm": 1.624780535697937,
+      "learning_rate": 9.989889390177467e-06,
+      "loss": 1.3492,
+      "step": 130
+    },
+    {
+      "epoch": 0.053903166811050146,
+      "grad_norm": 1.5940591096878052,
+      "learning_rate": 9.985537824862647e-06,
+      "loss": 1.2331,
+      "step": 140
+    },
+    {
+      "epoch": 0.057753393011839446,
+      "grad_norm": 1.2809606790542603,
+      "learning_rate": 9.980410816696317e-06,
+      "loss": 1.1401,
+      "step": 150
+    },
+    {
+      "epoch": 0.06160361921262874,
+      "grad_norm": 1.2885082960128784,
+      "learning_rate": 9.97450916312541e-06,
+      "loss": 1.058,
+      "step": 160
+    },
+    {
+      "epoch": 0.06545384541341803,
+      "grad_norm": 0.8517888188362122,
+      "learning_rate": 9.967833782084004e-06,
+      "loss": 0.9987,
+      "step": 170
+    },
+    {
+      "epoch": 0.06930407161420733,
+      "grad_norm": 0.7512567043304443,
+      "learning_rate": 9.960385711850554e-06,
+      "loss": 0.9651,
+      "step": 180
+    },
+    {
+      "epoch": 0.07315429781499663,
+      "grad_norm": 0.5244823098182678,
+      "learning_rate": 9.952166110886394e-06,
+      "loss": 0.9371,
+      "step": 190
+    },
+    {
+      "epoch": 0.07700452401578593,
+      "grad_norm": 0.5010038614273071,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.9026,
+      "step": 200
+    },
+    {
+      "epoch": 0.08085475021657522,
+      "grad_norm": 0.4211359918117523,
+      "learning_rate": 9.933417550425952e-06,
+      "loss": 0.8992,
+      "step": 210
+    },
+    {
+      "epoch": 0.08470497641736452,
+      "grad_norm": 0.46307405829429626,
+      "learning_rate": 9.9228915070518e-06,
+      "loss": 0.8775,
+      "step": 220
+    },
+    {
+      "epoch": 0.08855520261815382,
+      "grad_norm": 0.3602840006351471,
+      "learning_rate": 9.911599764737632e-06,
+      "loss": 0.8684,
+      "step": 230
+    },
+    {
+      "epoch": 0.09240542881894312,
+      "grad_norm": 0.40888720750808716,
+      "learning_rate": 9.899544079783606e-06,
+      "loss": 0.8526,
+      "step": 240
+    },
+    {
+      "epoch": 0.0962556550197324,
+      "grad_norm": 0.40349289774894714,
+      "learning_rate": 9.886726327312338e-06,
+      "loss": 0.8673,
+      "step": 250
+    },
+    {
+      "epoch": 0.1001058812205217,
+      "grad_norm": 0.41821858286857605,
+      "learning_rate": 9.87314850097724e-06,
+      "loss": 0.8586,
+      "step": 260
+    },
+    {
+      "epoch": 0.103956107421311,
+      "grad_norm": 0.43613502383232117,
+      "learning_rate": 9.858812712652452e-06,
+      "loss": 0.8343,
+      "step": 270
+    },
+    {
+      "epoch": 0.10780633362210029,
+      "grad_norm": 0.4173331558704376,
+      "learning_rate": 9.843721192104341e-06,
+      "loss": 0.8282,
+      "step": 280
+    },
+    {
+      "epoch": 0.11165655982288959,
+      "grad_norm": 0.43987053632736206,
+      "learning_rate": 9.827876286644698e-06,
+      "loss": 0.8289,
+      "step": 290
+    },
+    {
+      "epoch": 0.11550678602367889,
+      "grad_norm": 0.44678017497062683,
+      "learning_rate": 9.811280460765643e-06,
+      "loss": 0.7853,
+      "step": 300
+    },
+    {
+      "epoch": 0.11935701222446819,
+      "grad_norm": 0.42063668370246887,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.7865,
+      "step": 310
+    },
+    {
+      "epoch": 0.12320723842525748,
+      "grad_norm": 0.39193791151046753,
+      "learning_rate": 9.775846489301279e-06,
+      "loss": 0.7783,
+      "step": 320
+    },
+    {
+      "epoch": 0.12705746462604678,
+      "grad_norm": 0.3716551959514618,
+      "learning_rate": 9.757013855061157e-06,
+      "loss": 0.7711,
+      "step": 330
+    },
+    {
+      "epoch": 0.13090769082683607,
+      "grad_norm": 0.33533889055252075,
+      "learning_rate": 9.737441322234758e-06,
+      "loss": 0.7721,
+      "step": 340
+    },
+    {
+      "epoch": 0.13475791702762538,
+      "grad_norm": 0.33642351627349854,
+      "learning_rate": 9.717131935103605e-06,
+      "loss": 0.7728,
+      "step": 350
+    },
+    {
+      "epoch": 0.13860814322841467,
+      "grad_norm": 0.3166275918483734,
+      "learning_rate": 9.696088852558398e-06,
+      "loss": 0.7558,
+      "step": 360
+    },
+    {
+      "epoch": 0.14245836942920398,
+      "grad_norm": 0.2903101444244385,
+      "learning_rate": 9.674315347607686e-06,
+      "loss": 0.775,
+      "step": 370
+    },
+    {
+      "epoch": 0.14630859562999327,
+      "grad_norm": 0.2688869535923004,
+      "learning_rate": 9.651814806868796e-06,
+      "loss": 0.7458,
+      "step": 380
+    },
+    {
+      "epoch": 0.15015882183078255,
+      "grad_norm": 0.34237435460090637,
+      "learning_rate": 9.62859073004107e-06,
+      "loss": 0.7436,
+      "step": 390
+    },
+    {
+      "epoch": 0.15400904803157187,
+      "grad_norm": 0.2757215201854706,
+      "learning_rate": 9.604646729361547e-06,
+      "loss": 0.742,
+      "step": 400
+    },
+    {
+      "epoch": 0.15785927423236115,
+      "grad_norm": 0.26144126057624817,
+      "learning_rate": 9.579986529043099e-06,
+      "loss": 0.7238,
+      "step": 410
+    },
+    {
+      "epoch": 0.16170950043315044,
+      "grad_norm": 0.2591111958026886,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.7412,
+      "step": 420
+    },
+    {
+      "epoch": 0.16555972663393975,
+      "grad_norm": 0.26062995195388794,
+      "learning_rate": 9.528532982727281e-06,
+      "loss": 0.7325,
+      "step": 430
+    },
+    {
+      "epoch": 0.16940995283472904,
+      "grad_norm": 0.25959473848342896,
+      "learning_rate": 9.501747639735018e-06,
+      "loss": 0.7263,
+      "step": 440
+    },
+    {
+      "epoch": 0.17326017903551832,
+      "grad_norm": 0.2505236566066742,
+      "learning_rate": 9.474262101869272e-06,
+      "loss": 0.7156,
+      "step": 450
+    },
+    {
+      "epoch": 0.17711040523630764,
+      "grad_norm": 0.24393941462039948,
+      "learning_rate": 9.446080644188141e-06,
+      "loss": 0.7287,
+      "step": 460
+    },
+    {
+      "epoch": 0.18096063143709692,
+      "grad_norm": 0.22590100765228271,
+      "learning_rate": 9.417207649992017e-06,
+      "loss": 0.7294,
+      "step": 470
+    },
+    {
+      "epoch": 0.18481085763788624,
+      "grad_norm": 0.2169860303401947,
+      "learning_rate": 9.387647610141816e-06,
+      "loss": 0.7205,
+      "step": 480
+    },
+    {
+      "epoch": 0.18866108383867553,
+      "grad_norm": 0.20913545787334442,
+      "learning_rate": 9.35740512236047e-06,
+      "loss": 0.7155,
+      "step": 490
+    },
+    {
+      "epoch": 0.1925113100394648,
+      "grad_norm": 0.23903530836105347,
+      "learning_rate": 9.326484890517809e-06,
+      "loss": 0.7199,
+      "step": 500
+    },
+    {
+      "epoch": 0.19636153624025413,
+      "grad_norm": 0.22560527920722961,
+      "learning_rate": 9.294891723898926e-06,
+      "loss": 0.7059,
+      "step": 510
+    },
+    {
+      "epoch": 0.2002117624410434,
+      "grad_norm": 0.23946906626224518,
+      "learning_rate": 9.262630536456153e-06,
+      "loss": 0.7405,
+      "step": 520
+    },
+    {
+      "epoch": 0.2040619886418327,
+      "grad_norm": 0.2219129353761673,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.7112,
+      "step": 530
+    },
+    {
+      "epoch": 0.207912214842622,
+      "grad_norm": 0.22051450610160828,
+      "learning_rate": 9.196124273642422e-06,
+      "loss": 0.7262,
+      "step": 540
+    },
+    {
+      "epoch": 0.2117624410434113,
+      "grad_norm": 0.18761171400547028,
+      "learning_rate": 9.161889542552842e-06,
+      "loss": 0.7367,
+      "step": 550
+    },
+    {
+      "epoch": 0.21561266724420058,
+      "grad_norm": 0.20020392537117004,
+      "learning_rate": 9.127007477593193e-06,
+      "loss": 0.7304,
+      "step": 560
+    },
+    {
+      "epoch": 0.2194628934449899,
+      "grad_norm": 0.3467169404029846,
+      "learning_rate": 9.091483504265974e-06,
+      "loss": 0.7354,
+      "step": 570
+    },
+    {
+      "epoch": 0.22331311964577918,
+      "grad_norm": 0.20516636967658997,
+      "learning_rate": 9.055323147915116e-06,
+      "loss": 0.7138,
+      "step": 580
+    },
+    {
+      "epoch": 0.2271633458465685,
+      "grad_norm": 0.21599572896957397,
+      "learning_rate": 9.018532032866585e-06,
+      "loss": 0.7277,
+      "step": 590
+    },
+    {
+      "epoch": 0.23101357204735778,
+      "grad_norm": 0.19692525267601013,
+      "learning_rate": 8.981115881553579e-06,
+      "loss": 0.726,
+      "step": 600
+    },
+    {
+      "epoch": 0.23486379824814707,
+      "grad_norm": 0.21675938367843628,
+      "learning_rate": 8.943080513626476e-06,
+      "loss": 0.7355,
+      "step": 610
+    },
+    {
+      "epoch": 0.23871402444893638,
+      "grad_norm": 0.1960447132587433,
+      "learning_rate": 8.904431845047641e-06,
+      "loss": 0.7266,
+      "step": 620
+    },
+    {
+      "epoch": 0.24256425064972567,
+      "grad_norm": 0.20055189728736877,
+      "learning_rate": 8.86517588717128e-06,
+      "loss": 0.733,
+      "step": 630
+    },
+    {
+      "epoch": 0.24641447685051496,
+      "grad_norm": 0.19034889340400696,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.7319,
+      "step": 640
+    },
+    {
+      "epoch": 0.25026470305130427,
+      "grad_norm": 0.18107040226459503,
+      "learning_rate": 8.784866620277314e-06,
+      "loss": 0.6893,
+      "step": 650
+    },
+    {
+      "epoch": 0.25411492925209356,
+      "grad_norm": 0.21649578213691711,
+      "learning_rate": 8.743825802439018e-06,
+      "loss": 0.7042,
+      "step": 660
+    },
+    {
+      "epoch": 0.25796515545288284,
+      "grad_norm": 0.18575124442577362,
+      "learning_rate": 8.702202675718956e-06,
+      "loss": 0.7286,
+      "step": 670
+    },
+    {
+      "epoch": 0.26181538165367213,
+      "grad_norm": 0.19810175895690918,
+      "learning_rate": 8.660003714113959e-06,
+      "loss": 0.7007,
+      "step": 680
+    },
+    {
+      "epoch": 0.26566560785446147,
+      "grad_norm": 0.1770690232515335,
+      "learning_rate": 8.617235481185323e-06,
+      "loss": 0.7129,
+      "step": 690
+    },
+    {
+      "epoch": 0.26951583405525076,
+      "grad_norm": 0.19811353087425232,
+      "learning_rate": 8.573904629037931e-06,
+      "loss": 0.7213,
+      "step": 700
+    },
+    {
+      "epoch": 0.27336606025604004,
+      "grad_norm": 0.19989052414894104,
+      "learning_rate": 8.530017897285585e-06,
+      "loss": 0.7046,
+      "step": 710
+    },
+    {
+      "epoch": 0.27721628645682933,
+      "grad_norm": 0.1774475872516632,
+      "learning_rate": 8.485582112002746e-06,
+      "loss": 0.7156,
+      "step": 720
+    },
+    {
+      "epoch": 0.2810665126576186,
+      "grad_norm": 0.18944570422172546,
+      "learning_rate": 8.440604184662811e-06,
+      "loss": 0.7146,
+      "step": 730
+    },
+    {
+      "epoch": 0.28491673885840796,
+      "grad_norm": 0.20754915475845337,
+      "learning_rate": 8.395091111063118e-06,
+      "loss": 0.7019,
+      "step": 740
+    },
+    {
+      "epoch": 0.28876696505919724,
+      "grad_norm": 0.18284611403942108,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.7024,
+      "step": 750
+    },
+    {
+      "epoch": 0.29261719125998653,
+      "grad_norm": 0.1854391247034073,
+      "learning_rate": 8.302487923351842e-06,
+      "loss": 0.6921,
+      "step": 760
+    },
+    {
+      "epoch": 0.2964674174607758,
+      "grad_norm": 0.17971260845661163,
+      "learning_rate": 8.255412212597021e-06,
+      "loss": 0.7106,
+      "step": 770
+    },
+    {
+      "epoch": 0.3003176436615651,
+      "grad_norm": 0.1907491832971573,
+      "learning_rate": 8.207830160055688e-06,
+      "loss": 0.7077,
+      "step": 780
+    },
+    {
+      "epoch": 0.3041678698623544,
+      "grad_norm": 0.18174287676811218,
+      "learning_rate": 8.159749166566786e-06,
+      "loss": 0.6935,
+      "step": 790
+    },
+    {
+      "epoch": 0.30801809606314373,
+      "grad_norm": 0.18525435030460358,
+      "learning_rate": 8.11117671057377e-06,
+      "loss": 0.7173,
+      "step": 800
+    },
+    {
+      "epoch": 0.311868322263933,
+      "grad_norm": 0.1778651922941208,
+      "learning_rate": 8.062120346961404e-06,
+      "loss": 0.7031,
+      "step": 810
+    },
+    {
+      "epoch": 0.3157185484647223,
+      "grad_norm": 0.16080085933208466,
+      "learning_rate": 8.012587705880705e-06,
+      "loss": 0.7079,
+      "step": 820
+    },
+    {
+      "epoch": 0.3195687746655116,
+      "grad_norm": 0.19747675955295563,
+      "learning_rate": 7.962586491562146e-06,
+      "loss": 0.6968,
+      "step": 830
+    },
+    {
+      "epoch": 0.3234190008663009,
+      "grad_norm": 0.188969224691391,
+      "learning_rate": 7.912124481117357e-06,
+      "loss": 0.6977,
+      "step": 840
+    },
+    {
+      "epoch": 0.3272692270670902,
+      "grad_norm": 0.17062774300575256,
+      "learning_rate": 7.861209523329494e-06,
+      "loss": 0.7046,
+      "step": 850
+    },
+    {
+      "epoch": 0.3311194532678795,
+      "grad_norm": 0.19438420236110687,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.6947,
+      "step": 860
+    },
+    {
+      "epoch": 0.3349696794686688,
+      "grad_norm": 0.18522900342941284,
+      "learning_rate": 7.758052511879035e-06,
+      "loss": 0.6917,
+      "step": 870
+    },
+    {
+      "epoch": 0.3388199056694581,
+      "grad_norm": 0.16711556911468506,
+      "learning_rate": 7.705826503098642e-06,
+      "loss": 0.6988,
+      "step": 880
+    },
+    {
+      "epoch": 0.34267013187024736,
+      "grad_norm": 0.17881284654140472,
+      "learning_rate": 7.653179634243979e-06,
+      "loss": 0.6953,
+      "step": 890
+    },
+    {
+      "epoch": 0.34652035807103665,
+      "grad_norm": 0.16910958290100098,
+      "learning_rate": 7.600120093927696e-06,
+      "loss": 0.6882,
+      "step": 900
+    },
+    {
+      "epoch": 0.350370584271826,
+      "grad_norm": 0.19626165926456451,
+      "learning_rate": 7.546656134948729e-06,
+      "loss": 0.6871,
+      "step": 910
+    },
+    {
+      "epoch": 0.3542208104726153,
+      "grad_norm": 0.18360614776611328,
+      "learning_rate": 7.492796073008661e-06,
+      "loss": 0.684,
+      "step": 920
+    },
+    {
+      "epoch": 0.35807103667340456,
+      "grad_norm": 0.18477146327495575,
+      "learning_rate": 7.438548285418326e-06,
+      "loss": 0.6786,
+      "step": 930
+    },
+    {
+      "epoch": 0.36192126287419385,
+      "grad_norm": 0.17814669013023376,
+      "learning_rate": 7.3839212097948e-06,
+      "loss": 0.7134,
+      "step": 940
+    },
+    {
+      "epoch": 0.36577148907498314,
+      "grad_norm": 0.16895264387130737,
+      "learning_rate": 7.328923342749039e-06,
+      "loss": 0.7255,
+      "step": 950
+    },
+    {
+      "epoch": 0.3696217152757725,
+      "grad_norm": 0.38183388113975525,
+      "learning_rate": 7.273563238564319e-06,
+      "loss": 0.7036,
+      "step": 960
+    },
+    {
+      "epoch": 0.37347194147656176,
+      "grad_norm": 0.1762356013059616,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.6823,
+      "step": 970
+    },
+    {
+      "epoch": 0.37732216767735105,
+      "grad_norm": 0.16576291620731354,
+      "learning_rate": 7.161790816280858e-06,
+      "loss": 0.697,
+      "step": 980
+    },
+    {
+      "epoch": 0.38117239387814034,
+      "grad_norm": 0.18940085172653198,
+      "learning_rate": 7.1053958830920125e-06,
+      "loss": 0.6712,
+      "step": 990
+    },
+    {
+      "epoch": 0.3850226200789296,
+      "grad_norm": 0.1592113971710205,
+      "learning_rate": 7.0486734798799715e-06,
+      "loss": 0.7022,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3888728462797189,
+      "grad_norm": 0.1830289512872696,
+      "learning_rate": 6.991632429159708e-06,
+      "loss": 0.7227,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39272307248050825,
+      "grad_norm": 0.18130671977996826,
+      "learning_rate": 6.934281603008132e-06,
+      "loss": 0.6963,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39657329868129754,
+      "grad_norm": 0.17706353962421417,
+      "learning_rate": 6.876629921684143e-06,
+      "loss": 0.7089,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4004235248820868,
+      "grad_norm": 0.18082036077976227,
+      "learning_rate": 6.818686352241199e-06,
+      "loss": 0.706,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4042737510828761,
+      "grad_norm": 0.17183524370193481,
+      "learning_rate": 6.760459907132573e-06,
+      "loss": 0.6932,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4081239772836654,
+      "grad_norm": 0.17395102977752686,
+      "learning_rate": 6.701959642809589e-06,
+      "loss": 0.6968,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41197420348445474,
+      "grad_norm": 0.17546795308589935,
+      "learning_rate": 6.643194658312987e-06,
+      "loss": 0.6813,
+      "step": 1070
+    },
+    {
+      "epoch": 0.415824429685244,
+      "grad_norm": 0.1862875074148178,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.7186,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4196746558860333,
+      "grad_norm": 0.188532754778862,
+      "learning_rate": 6.5249071294110735e-06,
+      "loss": 0.6915,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4235248820868226,
+      "grad_norm": 0.17434121668338776,
+      "learning_rate": 6.465402983265274e-06,
+      "loss": 0.7085,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4273751082876119,
+      "grad_norm": 0.16904865205287933,
+      "learning_rate": 6.405670910603243e-06,
+      "loss": 0.7057,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43122533448840117,
+      "grad_norm": 0.1883309930562973,
+      "learning_rate": 6.345720202059291e-06,
+      "loss": 0.6974,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4350755606891905,
+      "grad_norm": 0.1909874975681305,
+      "learning_rate": 6.285560182274008e-06,
+      "loss": 0.688,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4389257868899798,
+      "grad_norm": 0.17202404141426086,
+      "learning_rate": 6.225200208443936e-06,
+      "loss": 0.6937,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4427760130907691,
+      "grad_norm": 0.16120585799217224,
+      "learning_rate": 6.16464966886616e-06,
+      "loss": 0.687,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44662623929155837,
+      "grad_norm": 0.19506746530532837,
+      "learning_rate": 6.103917981478065e-06,
+      "loss": 0.7154,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45047646549234766,
+      "grad_norm": 0.17102709412574768,
+      "learning_rate": 6.043014592392489e-06,
+      "loss": 0.721,
+      "step": 1170
+    },
+    {
+      "epoch": 0.454326691693137,
+      "grad_norm": 0.16543884575366974,
+      "learning_rate": 5.981948974428485e-06,
+      "loss": 0.6957,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4581769178939263,
+      "grad_norm": 0.18185384571552277,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.6741,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46202714409471557,
+      "grad_norm": 0.21489648520946503,
+      "learning_rate": 5.8593690678282355e-06,
+      "loss": 0.6897,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46587737029550486,
+      "grad_norm": 0.18276765942573547,
+      "learning_rate": 5.797873845081301e-06,
+      "loss": 0.6901,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46972759649629414,
+      "grad_norm": 0.1854240745306015,
+      "learning_rate": 5.736254522269075e-06,
+      "loss": 0.7034,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47357782269708343,
+      "grad_norm": 0.17562611401081085,
+      "learning_rate": 5.6745206835658386e-06,
+      "loss": 0.7083,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47742804889787277,
+      "grad_norm": 0.16679899394512177,
+      "learning_rate": 5.612681930957495e-06,
+      "loss": 0.6864,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48127827509866206,
+      "grad_norm": 0.16328389942646027,
+      "learning_rate": 5.550747882748095e-06,
+      "loss": 0.6803,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48512850129945134,
+      "grad_norm": 0.16674289107322693,
+      "learning_rate": 5.4887281720638195e-06,
+      "loss": 0.6859,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48897872750024063,
+      "grad_norm": 0.1864020675420761,
+      "learning_rate": 5.42663244535466e-06,
+      "loss": 0.6759,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4928289537010299,
+      "grad_norm": 0.172596275806427,
+      "learning_rate": 5.3644703608940205e-06,
+      "loss": 0.6802,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49667917990181926,
+      "grad_norm": 0.17047473788261414,
+      "learning_rate": 5.302251587276491e-06,
+      "loss": 0.7002,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5005294061026085,
+      "grad_norm": 0.17811772227287292,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.6952,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5043796323033978,
+      "grad_norm": 0.18068239092826843,
+      "learning_rate": 5.177682689530615e-06,
+      "loss": 0.7096,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5082298585041871,
+      "grad_norm": 0.1720874309539795,
+      "learning_rate": 5.11535194065619e-06,
+      "loss": 0.6815,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5120800847049765,
+      "grad_norm": 0.18174390494823456,
+      "learning_rate": 5.053003250119113e-06,
+      "loss": 0.6822,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5159303109057657,
+      "grad_norm": 0.1736256331205368,
+      "learning_rate": 4.9906463155383925e-06,
+      "loss": 0.6931,
+      "step": 1340
+    },
+    {
+      "epoch": 0.519780537106555,
+      "grad_norm": 0.16942450404167175,
+      "learning_rate": 4.9282908358153046e-06,
+      "loss": 0.6896,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5236307633073443,
+      "grad_norm": 0.190751850605011,
+      "learning_rate": 4.8659465096248345e-06,
+      "loss": 0.6929,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5274809895081336,
+      "grad_norm": 0.17848949134349823,
+      "learning_rate": 4.803623033907169e-06,
+      "loss": 0.6961,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5313312157089229,
+      "grad_norm": 0.16205978393554688,
+      "learning_rate": 4.741330102359441e-06,
+      "loss": 0.7052,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5351814419097122,
+      "grad_norm": 0.1789245307445526,
+      "learning_rate": 4.679077403927996e-06,
+      "loss": 0.6789,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5390316681105015,
+      "grad_norm": 0.17830564081668854,
+      "learning_rate": 4.616874621301373e-06,
+      "loss": 0.689,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5428818943112907,
+      "grad_norm": 0.18352821469306946,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.7028,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5467321205120801,
+      "grad_norm": 0.17873211205005646,
+      "learning_rate": 4.492657493892822e-06,
+      "loss": 0.6833,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5505823467128694,
+      "grad_norm": 0.17903704941272736,
+      "learning_rate": 4.430662469650999e-06,
+      "loss": 0.6966,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5544325729136587,
+      "grad_norm": 0.18541929125785828,
+      "learning_rate": 4.368755999289118e-06,
+      "loss": 0.7,
+      "step": 1440
+    },
+    {
+      "epoch": 0.558282799114448,
+      "grad_norm": 0.18502700328826904,
+      "learning_rate": 4.306947711643949e-06,
+      "loss": 0.6832,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5621330253152372,
+      "grad_norm": 0.1658344566822052,
+      "learning_rate": 4.245247220281068e-06,
+      "loss": 0.6903,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5659832515160266,
+      "grad_norm": 0.16841623187065125,
+      "learning_rate": 4.183664121999587e-06,
+      "loss": 0.6854,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5698334777168159,
+      "grad_norm": 0.175063356757164,
+      "learning_rate": 4.122207995339474e-06,
+      "loss": 0.6828,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5736837039176051,
+      "grad_norm": 0.17930425703525543,
+      "learning_rate": 4.060888399091734e-06,
+      "loss": 0.6897,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5775339301183945,
+      "grad_norm": 0.1777365803718567,
+      "learning_rate": 3.999714870811639e-06,
+      "loss": 0.7025,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5813841563191837,
+      "grad_norm": 0.17292818427085876,
+      "learning_rate": 3.938696925335275e-06,
+      "loss": 0.6928,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5852343825199731,
+      "grad_norm": 0.17773698270320892,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.6965,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5890846087207623,
+      "grad_norm": 0.18144063651561737,
+      "learning_rate": 3.817165719666382e-06,
+      "loss": 0.6942,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5929348349215516,
+      "grad_norm": 0.1659162938594818,
+      "learning_rate": 3.756671362249841e-06,
+      "loss": 0.7048,
+      "step": 1540
+    },
+    {
+      "epoch": 0.596785061122341,
+      "grad_norm": 0.1686103194952011,
+      "learning_rate": 3.6963703902488907e-06,
+      "loss": 0.6952,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6006352873231302,
+      "grad_norm": 0.1856197565793991,
+      "learning_rate": 3.6362721827835655e-06,
+      "loss": 0.7026,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6044855135239195,
+      "grad_norm": 0.18449465930461884,
+      "learning_rate": 3.576386087436212e-06,
+      "loss": 0.6972,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6083357397247088,
+      "grad_norm": 0.17248596251010895,
+      "learning_rate": 3.5167214187975913e-06,
+      "loss": 0.6995,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6121859659254981,
+      "grad_norm": 0.15883755683898926,
+      "learning_rate": 3.457287457018086e-06,
+      "loss": 0.6936,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6160361921262875,
+      "grad_norm": 0.1613185554742813,
+      "learning_rate": 3.3980934463643003e-06,
+      "loss": 0.683,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6198864183270767,
+      "grad_norm": 0.1591046005487442,
+      "learning_rate": 3.3391485937812053e-06,
+      "loss": 0.6614,
+      "step": 1610
+    },
+    {
+      "epoch": 0.623736644527866,
+      "grad_norm": 0.16835181415081024,
+      "learning_rate": 3.280462067460115e-06,
+      "loss": 0.6659,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6275868707286553,
+      "grad_norm": 0.15480470657348633,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.7011,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6314370969294446,
+      "grad_norm": 0.15441040694713593,
+      "learning_rate": 3.163900464051089e-06,
+      "loss": 0.6845,
+      "step": 1640
+    },
+    {
+      "epoch": 0.635287323130234,
+      "grad_norm": 0.17166781425476074,
+      "learning_rate": 3.1060435167748826e-06,
+      "loss": 0.6942,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6391375493310232,
+      "grad_norm": 0.189008429646492,
+      "learning_rate": 3.048481152564253e-06,
+      "loss": 0.6815,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6429877755318125,
+      "grad_norm": 0.19025088846683502,
+      "learning_rate": 2.9912223245804017e-06,
+      "loss": 0.7054,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6468380017326018,
+      "grad_norm": 0.16954927146434784,
+      "learning_rate": 2.934275938772981e-06,
+      "loss": 0.684,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6506882279333911,
+      "grad_norm": 0.17913800477981567,
+      "learning_rate": 2.8776508524948676e-06,
+      "loss": 0.6755,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6545384541341804,
+      "grad_norm": 0.16214677691459656,
+      "learning_rate": 2.821355873124503e-06,
+      "loss": 0.6743,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6583886803349697,
+      "grad_norm": 0.17050673067569733,
+      "learning_rate": 2.7653997566960144e-06,
+      "loss": 0.6943,
+      "step": 1710
+    },
+    {
+      "epoch": 0.662238906535759,
+      "grad_norm": 0.17057165503501892,
+      "learning_rate": 2.709791206537309e-06,
+      "loss": 0.6848,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6660891327365482,
+      "grad_norm": 0.16865073144435883,
+      "learning_rate": 2.6545388719163747e-06,
+      "loss": 0.6802,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6699393589373376,
+      "grad_norm": 0.1651163250207901,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.6928,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6737895851381268,
+      "grad_norm": 0.1662459522485733,
+      "learning_rate": 2.545137167997007e-06,
+      "loss": 0.6871,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6776398113389162,
+      "grad_norm": 0.1758705973625183,
+      "learning_rate": 2.4910048148705972e-06,
+      "loss": 0.6733,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6814900375397055,
+      "grad_norm": 0.18139290809631348,
+      "learning_rate": 2.4372627069793327e-06,
+      "loss": 0.7153,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6853402637404947,
+      "grad_norm": 0.17099013924598694,
+      "learning_rate": 2.3839192032876486e-06,
+      "loss": 0.6982,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6891904899412841,
+      "grad_norm": 0.16996456682682037,
+      "learning_rate": 2.3309826007617047e-06,
+      "loss": 0.7096,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6930407161420733,
+      "grad_norm": 0.16759929060935974,
+      "learning_rate": 2.278461133078881e-06,
+      "loss": 0.6756,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6968909423428626,
+      "grad_norm": 0.17027722299098969,
+      "learning_rate": 2.2263629693471258e-06,
+      "loss": 0.6696,
+      "step": 1810
+    },
+    {
+      "epoch": 0.700741168543652,
+      "grad_norm": 0.1870974600315094,
+      "learning_rate": 2.1746962128343414e-06,
+      "loss": 0.6926,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7045913947444412,
+      "grad_norm": 0.16645511984825134,
+      "learning_rate": 2.1234688997080183e-06,
+      "loss": 0.6897,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7084416209452306,
+      "grad_norm": 0.1854025274515152,
+      "learning_rate": 2.0726889977853053e-06,
+      "loss": 0.6868,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7122918471460198,
+      "grad_norm": 0.18024225533008575,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.6991,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7161420733468091,
+      "grad_norm": 0.1634608954191208,
+      "learning_rate": 1.9725029496425814e-06,
+      "loss": 0.6941,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7199922995475985,
+      "grad_norm": 0.15702742338180542,
+      "learning_rate": 1.923112386205727e-06,
+      "loss": 0.6895,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7238425257483877,
+      "grad_norm": 0.16656793653964996,
+      "learning_rate": 1.874200397115079e-06,
+      "loss": 0.6968,
+      "step": 1880
+    },
+    {
+      "epoch": 0.727692751949177,
+      "grad_norm": 0.17513522505760193,
+      "learning_rate": 1.825774590065857e-06,
+      "loss": 0.6836,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7315429781499663,
+      "grad_norm": 0.1735902577638626,
+      "learning_rate": 1.7778424971332813e-06,
+      "loss": 0.6898,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7353932043507556,
+      "grad_norm": 0.16840334236621857,
+      "learning_rate": 1.7304115736010407e-06,
+      "loss": 0.693,
+      "step": 1910
+    },
+    {
+      "epoch": 0.739243430551545,
+      "grad_norm": 0.17388713359832764,
+      "learning_rate": 1.6834891968017108e-06,
+      "loss": 0.6842,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7430936567523342,
+      "grad_norm": 0.1762254238128662,
+      "learning_rate": 1.6370826649692884e-06,
+      "loss": 0.6932,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7469438829531235,
+      "grad_norm": 0.17363286018371582,
+      "learning_rate": 1.5911991961040463e-06,
+      "loss": 0.707,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7507941091539128,
+      "grad_norm": 0.17828553915023804,
+      "learning_rate": 1.545845926849851e-06,
+      "loss": 0.6929,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7546443353547021,
+      "grad_norm": 0.16877488791942596,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.7083,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7584945615554913,
+      "grad_norm": 0.17205101251602173,
+      "learning_rate": 1.4567581203207238e-06,
+      "loss": 0.7065,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7623447877562807,
+      "grad_norm": 0.17017169296741486,
+      "learning_rate": 1.413037439625593e-06,
+      "loss": 0.6883,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76619501395707,
+      "grad_norm": 0.18157702684402466,
+      "learning_rate": 1.3698746695458847e-06,
+      "loss": 0.6906,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7700452401578592,
+      "grad_norm": 0.1562771052122116,
+      "learning_rate": 1.3272765235521745e-06,
+      "loss": 0.6908,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7738954663586486,
+      "grad_norm": 0.17113158106803894,
+      "learning_rate": 1.2852496272942877e-06,
+      "loss": 0.6912,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7777456925594378,
+      "grad_norm": 0.18750227987766266,
+      "learning_rate": 1.2438005175707396e-06,
+      "loss": 0.676,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7815959187602272,
+      "grad_norm": 0.1711566299200058,
+      "learning_rate": 1.2029356413120297e-06,
+      "loss": 0.6641,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7854461449610165,
+      "grad_norm": 0.1693297028541565,
+      "learning_rate": 1.1626613545778786e-06,
+      "loss": 0.6809,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7892963711618057,
+      "grad_norm": 0.16488952934741974,
+      "learning_rate": 1.1229839215686294e-06,
+      "loss": 0.6885,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7931465973625951,
+      "grad_norm": 0.1631680577993393,
+      "learning_rate": 1.0839095136509188e-06,
+      "loss": 0.6686,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7969968235633843,
+      "grad_norm": 0.1820286214351654,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.6787,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8008470497641736,
+      "grad_norm": 0.1814328134059906,
+      "learning_rate": 1.007593988643402e-06,
+      "loss": 0.67,
+      "step": 2080
+    },
+    {
+      "epoch": 0.804697275964963,
+      "grad_norm": 0.1738474816083908,
+      "learning_rate": 9.703647415524613e-07,
+      "loss": 0.6914,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8085475021657522,
+      "grad_norm": 0.16704653203487396,
+      "learning_rate": 9.337622577045485e-07,
+      "loss": 0.6864,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8123977283665416,
+      "grad_norm": 0.17539674043655396,
+      "learning_rate": 8.977922301934588e-07,
+      "loss": 0.6992,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8162479545673308,
+      "grad_norm": 0.16690489649772644,
+      "learning_rate": 8.624602537417009e-07,
+      "loss": 0.6976,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8200981807681201,
+      "grad_norm": 0.17107558250427246,
+      "learning_rate": 8.277718238303123e-07,
+      "loss": 0.6792,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8239484069689095,
+      "grad_norm": 0.17137207090854645,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.6915,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8277986331696987,
+      "grad_norm": 0.19430260360240936,
+      "learning_rate": 7.603470842324223e-07,
+      "loss": 0.6755,
+      "step": 2150
+    },
+    {
+      "epoch": 0.831648859370488,
+      "grad_norm": 0.1586805284023285,
+      "learning_rate": 7.276212616857575e-07,
+      "loss": 0.6927,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8354990855712773,
+      "grad_norm": 0.16295920312404633,
+      "learning_rate": 6.955599583279837e-07,
+      "loss": 0.6932,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8393493117720666,
+      "grad_norm": 0.16323770582675934,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.686,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8431995379728559,
+      "grad_norm": 0.1712789535522461,
+      "learning_rate": 6.334507521075711e-07,
+      "loss": 0.6956,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8470497641736452,
+      "grad_norm": 0.18739785254001617,
+      "learning_rate": 6.034125096149179e-07,
+      "loss": 0.6822,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8508999903744345,
+      "grad_norm": 0.1786557286977768,
+      "learning_rate": 5.740581055485722e-07,
+      "loss": 0.6678,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8547502165752238,
+      "grad_norm": 0.15127992630004883,
+      "learning_rate": 5.453921056471961e-07,
+      "loss": 0.6909,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8586004427760131,
+      "grad_norm": 0.1755937933921814,
+      "learning_rate": 5.174189685761288e-07,
+      "loss": 0.6828,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8624506689768023,
+      "grad_norm": 0.16850881278514862,
+      "learning_rate": 4.901430452338979e-07,
+      "loss": 0.7001,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8663008951775917,
+      "grad_norm": 0.17501966655254364,
+      "learning_rate": 4.6356857807548286e-07,
+      "loss": 0.6822,
+      "step": 2250
+    },
+    {
+      "epoch": 0.870151121378381,
+      "grad_norm": 0.16359397768974304,
+      "learning_rate": 4.376997004524541e-07,
+      "loss": 0.6715,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8740013475791703,
+      "grad_norm": 0.17183807492256165,
+      "learning_rate": 4.12540435970073e-07,
+      "loss": 0.673,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8778515737799596,
+      "grad_norm": 0.1618465930223465,
+      "learning_rate": 3.880946978614703e-07,
+      "loss": 0.695,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8817017999807488,
+      "grad_norm": 0.16226373612880707,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.6906,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8855520261815382,
+      "grad_norm": 0.17027181386947632,
+      "learning_rate": 3.41358898202781e-07,
+      "loss": 0.6762,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8894022523823275,
+      "grad_norm": 0.16380780935287476,
+      "learning_rate": 3.190761058667741e-07,
+      "loss": 0.6786,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8932524785831167,
+      "grad_norm": 0.16705431044101715,
+      "learning_rate": 2.9752137720206463e-07,
+      "loss": 0.6861,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8971027047839061,
+      "grad_norm": 0.16693079471588135,
+      "learning_rate": 2.7669806479785067e-07,
+      "loss": 0.6975,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9009529309846953,
+      "grad_norm": 0.16823536157608032,
+      "learning_rate": 2.566094074799763e-07,
+      "loss": 0.6848,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9048031571854847,
+      "grad_norm": 0.17647452652454376,
+      "learning_rate": 2.3725852980716557e-07,
+      "loss": 0.6755,
+      "step": 2350
+    },
+    {
+      "epoch": 0.908653383386274,
+      "grad_norm": 0.16709265112876892,
+      "learning_rate": 2.1864844158503983e-07,
+      "loss": 0.686,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9125036095870632,
+      "grad_norm": 0.174675852060318,
+      "learning_rate": 2.007820373979702e-07,
+      "loss": 0.6745,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9163538357878526,
+      "grad_norm": 0.16284261643886566,
+      "learning_rate": 1.8366209615886532e-07,
+      "loss": 0.6804,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9202040619886418,
+      "grad_norm": 0.16767777502536774,
+      "learning_rate": 1.6729128067693613e-07,
+      "loss": 0.7027,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9240542881894311,
+      "grad_norm": 0.174700528383255,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.6912,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9279045143902205,
+      "grad_norm": 0.17113305628299713,
+      "learning_rate": 1.3680709523610058e-07,
+      "loss": 0.677,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9317547405910097,
+      "grad_norm": 0.1538524329662323,
+      "learning_rate": 1.2269846674030418e-07,
+      "loss": 0.7005,
+      "step": 2420
+    },
+    {
+      "epoch": 0.935604966791799,
+      "grad_norm": 0.17491945624351501,
+      "learning_rate": 1.0934844619042595e-07,
+      "loss": 0.6979,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9394551929925883,
+      "grad_norm": 0.1743723750114441,
+      "learning_rate": 9.675911002803961e-08,
+      "loss": 0.6837,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9433054191933776,
+      "grad_norm": 0.17521455883979797,
+      "learning_rate": 8.493241637904381e-08,
+      "loss": 0.6806,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9471556453941669,
+      "grad_norm": 0.16601204872131348,
+      "learning_rate": 7.387020474909689e-08,
+      "loss": 0.7012,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9510058715949562,
+      "grad_norm": 0.17316219210624695,
+      "learning_rate": 6.357419573750856e-08,
+      "loss": 0.6886,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9548560977957455,
+      "grad_norm": 0.16941070556640625,
+      "learning_rate": 5.4045990769613365e-08,
+      "loss": 0.6824,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9587063239965348,
+      "grad_norm": 0.16851216554641724,
+      "learning_rate": 4.5287071847694434e-08,
+      "loss": 0.6822,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9625565501973241,
+      "grad_norm": 0.16457901895046234,
+      "learning_rate": 3.7298801320467817e-08,
+      "loss": 0.6823,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9664067763981133,
+      "grad_norm": 0.15943965315818787,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.6753,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9702570025989027,
+      "grad_norm": 0.16553722321987152,
+      "learning_rate": 2.3639055324410355e-08,
+      "loss": 0.6952,
+      "step": 2520
+    },
+    {
+      "epoch": 0.974107228799692,
+      "grad_norm": 0.16341909766197205,
+      "learning_rate": 1.7969704471369477e-08,
+      "loss": 0.6797,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9779574550004813,
+      "grad_norm": 0.1563764363527298,
+      "learning_rate": 1.3075250914142413e-08,
+      "loss": 0.6847,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9818076812012706,
+      "grad_norm": 0.161655992269516,
+      "learning_rate": 8.956455928476049e-09,
+      "loss": 0.6938,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9856579074020598,
+      "grad_norm": 0.15775270760059357,
+      "learning_rate": 5.6139601453786275e-09,
+      "loss": 0.6755,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9895081336028492,
+      "grad_norm": 0.16670598089694977,
+      "learning_rate": 3.0482834514805602e-09,
+      "loss": 0.6791,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9933583598036385,
+      "grad_norm": 0.16223575174808502,
+      "learning_rate": 1.2598249081713366e-09,
+      "loss": 0.6632,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9972085860044277,
+      "grad_norm": 0.1705436408519745,
+      "learning_rate": 2.488626895302826e-10,
+      "loss": 0.6955,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9999037443449803,
+      "step": 2597,
+      "total_flos": 5.142881086048567e+18,
+      "train_loss": 0.7818009450337406,
+      "train_runtime": 16208.9036,
+      "train_samples_per_second": 41.021,
+      "train_steps_per_second": 0.16
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2597,
+  "num_input_tokens_seen": 332174164,
+  "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": 5.142881086048567e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3base/siglip2_g16_384/finetune/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9ef0bc1f9f699da273bdcc779f1762c3c72c4ab5154dc642a0dc6a794890e4a
+size 6264

continuous/qwen3base/siglip2_g16_384/pretrain/config.json

@@ -0,0 +1,89 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1536,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "flat",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_tunable_parts": "mm_mlp_adapter",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "open_clip_hub:ViT-gopt-16-SigLIP2-384",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "pad_token_id": 151643,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "transformers_version": "4.57.6",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": "/cache/ckpt/download/tokenizer/continuous/siglip2/siglip2_g16_384.npz",
+  "vocab_size": 151936
+}

continuous/qwen3base/siglip2_g16_384/pretrain/mm_projector.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2d1fe4c7546340ff374e37aa5b9c0c1e52f536e597fba8939f6a9b8ccaa50443
+size 29378616

continuous/qwen3base/siglip2_g16_384/pretrain/trainer_state.json

@@ -0,0 +1,421 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 545,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01834862385321101,
+      "grad_norm": 1.0132722854614258,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 7.0746,
+      "step": 10
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 0.17718914151191711,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 5.6697,
+      "step": 20
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.0761205330491066,
+      "learning_rate": 0.0009999907688176173,
+      "loss": 5.363,
+      "step": 30
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.07981567084789276,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 5.1928,
+      "step": 40
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.04914864897727966,
+      "learning_rate": 0.0009959345572940772,
+      "loss": 5.1013,
+      "step": 50
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.046560291200876236,
+      "learning_rate": 0.0009911550081155983,
+      "loss": 5.0553,
+      "step": 60
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.07636922597885132,
+      "learning_rate": 0.0009845624347566433,
+      "loss": 5.0016,
+      "step": 70
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.07950583845376968,
+      "learning_rate": 0.0009761811727013547,
+      "loss": 4.9752,
+      "step": 80
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.10644080489873886,
+      "learning_rate": 0.0009660421601067666,
+      "loss": 4.9199,
+      "step": 90
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.0751793310046196,
+      "learning_rate": 0.0009541828235992926,
+      "loss": 4.8668,
+      "step": 100
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.08953722566366196,
+      "learning_rate": 0.0009406469401200151,
+      "loss": 4.807,
+      "step": 110
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.08909693360328674,
+      "learning_rate": 0.0009254844753287492,
+      "loss": 4.7566,
+      "step": 120
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.07628384977579117,
+      "learning_rate": 0.0009087513991633924,
+      "loss": 4.6926,
+      "step": 130
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.08165223151445389,
+      "learning_rate": 0.0008905094792353917,
+      "loss": 4.6485,
+      "step": 140
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.07996184378862381,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 4.6151,
+      "step": 150
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.11854229122400284,
+      "learning_rate": 0.0008497737783108237,
+      "loss": 4.5777,
+      "step": 160
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.12007135152816772,
+      "learning_rate": 0.0008274303669726426,
+      "loss": 4.5485,
+      "step": 170
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.08624158799648285,
+      "learning_rate": 0.0008038782961218137,
+      "loss": 4.5242,
+      "step": 180
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.09420331567525864,
+      "learning_rate": 0.000779204504653894,
+      "loss": 4.4979,
+      "step": 190
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.06992446631193161,
+      "learning_rate": 0.0007535000721258791,
+      "loss": 4.4653,
+      "step": 200
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.0823693796992302,
+      "learning_rate": 0.00072685988254984,
+      "loss": 4.4379,
+      "step": 210
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.11635736376047134,
+      "learning_rate": 0.0006993822741429906,
+      "loss": 4.4274,
+      "step": 220
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.134722039103508,
+      "learning_rate": 0.0006711686763270818,
+      "loss": 4.3803,
+      "step": 230
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.0801408439874649,
+      "learning_rate": 0.0006423232353170831,
+      "loss": 4.3319,
+      "step": 240
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.08124314993619919,
+      "learning_rate": 0.0006129524296812335,
+      "loss": 4.2975,
+      "step": 250
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.07056912779808044,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 4.2873,
+      "step": 260
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.0889868438243866,
+      "learning_rate": 0.0005530699351157809,
+      "loss": 4.2605,
+      "step": 270
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.06625508517026901,
+      "learning_rate": 0.0005227792933277943,
+      "loss": 4.2461,
+      "step": 280
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.0703110322356224,
+      "learning_rate": 0.0004924045652352092,
+      "loss": 4.2396,
+      "step": 290
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.07424715906381607,
+      "learning_rate": 0.0004620578745374633,
+      "loss": 4.2315,
+      "step": 300
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.07583996653556824,
+      "learning_rate": 0.0004318512414382059,
+      "loss": 4.2411,
+      "step": 310
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.06999773532152176,
+      "learning_rate": 0.0004018961691397148,
+      "loss": 4.2305,
+      "step": 320
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.07908158004283905,
+      "learning_rate": 0.0003723032322457458,
+      "loss": 4.2167,
+      "step": 330
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.0994156152009964,
+      "learning_rate": 0.0003431816685921625,
+      "loss": 4.2086,
+      "step": 340
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.07243939489126205,
+      "learning_rate": 0.0003146389760120416,
+      "loss": 4.2115,
+      "step": 350
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.08027427643537521,
+      "learning_rate": 0.00028678051552373485,
+      "loss": 4.2152,
+      "step": 360
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.06935559958219528,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 4.1987,
+      "step": 370
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.06464925408363342,
+      "learning_rate": 0.0002335247266004629,
+      "loss": 4.207,
+      "step": 380
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.0682426393032074,
+      "learning_rate": 0.000208323983828828,
+      "loss": 4.1948,
+      "step": 390
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.06646565347909927,
+      "learning_rate": 0.0001841999188095224,
+      "loss": 4.1907,
+      "step": 400
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.060899898409843445,
+      "learning_rate": 0.00016124158186777676,
+      "loss": 4.1945,
+      "step": 410
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.05990604683756828,
+      "learning_rate": 0.0001395337202205211,
+      "loss": 4.1825,
+      "step": 420
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.06116066500544548,
+      "learning_rate": 0.00011915646514489292,
+      "loss": 4.189,
+      "step": 430
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.06339370459318161,
+      "learning_rate": 0.00010018503618578818,
+      "loss": 4.1905,
+      "step": 440
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.062215931713581085,
+      "learning_rate": 8.268946349432583e-05,
+      "loss": 4.1811,
+      "step": 450
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.06115243583917618,
+      "learning_rate": 6.673432932217117e-05,
+      "loss": 4.181,
+      "step": 460
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.05852242931723595,
+      "learning_rate": 5.2378529625954684e-05,
+      "loss": 4.1638,
+      "step": 470
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.06404488533735275,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 4.1719,
+      "step": 480
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.05810102820396423,
+      "learning_rate": 2.8670803372379538e-05,
+      "loss": 4.1716,
+      "step": 490
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.06114489585161209,
+      "learning_rate": 1.9406390288876586e-05,
+      "loss": 4.1826,
+      "step": 500
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.05932239815592766,
+      "learning_rate": 1.191601558631461e-05,
+      "loss": 4.1773,
+      "step": 510
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.057630591094493866,
+      "learning_rate": 6.2273288462559305e-06,
+      "loss": 4.1689,
+      "step": 520
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.06068297103047371,
+      "learning_rate": 2.361328992552314e-06,
+      "loss": 4.1717,
+      "step": 530
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.058271002024412155,
+      "learning_rate": 3.322867770044202e-07,
+      "loss": 4.1722,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "step": 545,
+      "total_flos": 4.938377282327675e+17,
+      "train_loss": 4.499262699512166,
+      "train_runtime": 5451.8337,
+      "train_samples_per_second": 102.374,
+      "train_steps_per_second": 0.1
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 545,
+  "num_input_tokens_seen": 31896544,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.938377282327675e+17,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/smollm2/clip_openai__l14/finetune/chat_template.jinja

@@ -0,0 +1,6 @@
+{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system
+You are a helpful AI assistant named SmolLM, trained by Hugging Face<|im_end|>
+' }}{% endif %}{{'<|im_start|>' + message['role'] + '
+' + message['content'] + '<|im_end|>' + '
+'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant
+' }}{% endif %}

continuous/smollm2/clip_openai__l14/finetune/config.json

@@ -0,0 +1,70 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaLlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "dtype": "bfloat16",
+  "eos_token_id": 2,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "head_dim": 64,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 8192,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "mm_hidden_size": 1024,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "flat",
+  "mm_projector_lr": 0.0001,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_tunable_parts": "mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "/cache/ckpt/download/tokenizer/continuous/clip-vit-large-patch14",
+  "mm_vision_tower_lr": null,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 32,
+  "pad_token_id": 0,
+  "pos_skipping_range": 4096,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 130000,
+  "tie_word_embeddings": true,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "transformers.js_config": {
+    "dtype": "q4",
+    "kv_cache_dtype": {
+      "fp16": "float16",
+      "q4f16": "float16"
+    },
+    "use_external_data_format": {
+      "model.onnx": true,
+      "model_fp16.onnx": true
+    }
+  },
+  "transformers_version": "4.57.6",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "vision_tower_pretrained": null,
+  "vocab_size": 49152
+}