Upload continuous evaluation checkpoints, batch 15

.gitattributes

@@ -142,3 +142,6 @@ continuous/qwen3/mc2_b32_224_mt5/finetune/tokenizer.json filter=lfs diff=lfs mer
 continuous/qwen3/mc2_b32_384/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 continuous/qwen3/mc2_g14_224/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 continuous/qwen3/mc2_g14_378/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+continuous/qwen3/mc2_h14_378/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+continuous/qwen3/mc2_l14_224/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+continuous/qwen3/mc2_m16_224/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text

continuous/qwen3/mc2_h14_378/finetune/model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b30fe8678b4c12b7482a433c9845f3f110f7c381b4968174f5418c4edd6e4bb8
+size 4733887200

continuous/qwen3/mc2_h14_378/finetune/model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3be8af601cdf5ad447cef4fd0d7f9348eab8f61f46fc6298c60a2f1ec80b6931
+size 622329984

continuous/qwen3/mc2_h14_378/finetune/model.safetensors.index.json

@@ -0,0 +1,714 @@
+{
+  "metadata": {
+    "total_size": 5356123648
+  },
+  "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-00001-of-00002.safetensors",
+    "model.mm_projector.0.weight": "model-00001-of-00002.safetensors",
+    "model.mm_projector.2.bias": "model-00001-of-00002.safetensors",
+    "model.mm_projector.2.weight": "model-00001-of-00002.safetensors",
+    "model.norm.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.class_embedding": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.conv1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.ln_post.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.ln_post.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.ln_pre.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.ln_pre.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.positional_embedding": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.proj": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.mlp.c_proj.weight": "model-00001-of-00002.safetensors"
+  }
+}

continuous/qwen3/mc2_h14_378/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/mc2_h14_378/finetune/tokenizer.json

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

continuous/qwen3/mc2_h14_378/finetune/tokenizer_config.json

@@ -0,0 +1,241 @@
+{
+  "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,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0].role == 'system' %}\n        {{- messages[0].content + '\\n\\n' }}\n    {%- endif %}\n    {{- \"# 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>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\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\" }}\n{%- else %}\n    {%- if messages[0].role == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0].content + '<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n    {%- set index = (messages|length - 1) - loop.index0 %}\n    {%- if ns.multi_step_tool and message.role == \"user\" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n        {%- set ns.multi_step_tool = false %}\n        {%- set ns.last_query_index = index %}\n    {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n    {%- if message.content is string %}\n        {%- set content = message.content %}\n    {%- else %}\n        {%- set content = '' %}\n    {%- endif %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set reasoning_content = '' %}\n        {%- if message.reasoning_content is string %}\n            {%- set reasoning_content = message.reasoning_content %}\n        {%- else %}\n            {%- if '</think>' in content %}\n                {%- set reasoning_content = content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n                {%- set content = content.split('</think>')[-1].lstrip('\\n') %}\n            {%- endif %}\n        {%- endif %}\n        {%- if loop.index0 > ns.last_query_index %}\n            {%- if loop.last or (not loop.last and reasoning_content) %}\n                {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n            {%- else %}\n                {{- '<|im_start|>' + message.role + '\\n' + content }}\n            {%- endif %}\n        {%- else %}\n            {{- '<|im_start|>' + message.role + '\\n' + content }}\n        {%- endif %}\n        {%- if message.tool_calls %}\n            {%- for tool_call in message.tool_calls %}\n                {%- if (loop.first and content) or (not loop.first) %}\n                    {{- '\\n' }}\n                {%- endif %}\n                {%- if tool_call.function %}\n                    {%- set tool_call = tool_call.function %}\n                {%- endif %}\n                {{- '<tool_call>\\n{\"name\": \"' }}\n                {{- tool_call.name }}\n                {{- '\", \"arguments\": ' }}\n                {%- if tool_call.arguments is string %}\n                    {{- tool_call.arguments }}\n                {%- else %}\n                    {{- tool_call.arguments | tojson }}\n                {%- endif %}\n                {{- '}\\n</tool_call>' }}\n            {%- endfor %}\n        {%- endif %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n    {%- if enable_thinking is defined and enable_thinking is false %}\n        {{- '<think>\\n\\n</think>\\n\\n' }}\n    {%- endif %}\n{%- endif %}",
+  "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/mc2_h14_378/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": 10.239805221557617,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 2.6496,
+      "step": 10
+    },
+    {
+      "epoch": 0.007700452401578592,
+      "grad_norm": 3.7262890338897705,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.922,
+      "step": 20
+    },
+    {
+      "epoch": 0.01155067860236789,
+      "grad_norm": 2.285139560699463,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.5233,
+      "step": 30
+    },
+    {
+      "epoch": 0.015400904803157185,
+      "grad_norm": 0.7836734056472778,
+      "learning_rate": 5e-06,
+      "loss": 1.1825,
+      "step": 40
+    },
+    {
+      "epoch": 0.019251131003946483,
+      "grad_norm": 0.48453113436698914,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.0671,
+      "step": 50
+    },
+    {
+      "epoch": 0.02310135720473578,
+      "grad_norm": 0.25541919469833374,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.9906,
+      "step": 60
+    },
+    {
+      "epoch": 0.026951583405525073,
+      "grad_norm": 0.20451484620571136,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.9485,
+      "step": 70
+    },
+    {
+      "epoch": 0.03080180960631437,
+      "grad_norm": 0.17761553823947906,
+      "learning_rate": 9.999996111488723e-06,
+      "loss": 0.9448,
+      "step": 80
+    },
+    {
+      "epoch": 0.034652035807103666,
+      "grad_norm": 0.19146940112113953,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 0.9172,
+      "step": 90
+    },
+    {
+      "epoch": 0.038502262007892966,
+      "grad_norm": 0.1735222190618515,
+      "learning_rate": 9.998285264324193e-06,
+      "loss": 0.8921,
+      "step": 100
+    },
+    {
+      "epoch": 0.04235248820868226,
+      "grad_norm": 0.17756135761737823,
+      "learning_rate": 9.996263605626052e-06,
+      "loss": 0.8686,
+      "step": 110
+    },
+    {
+      "epoch": 0.04620271440947156,
+      "grad_norm": 0.16893678903579712,
+      "learning_rate": 9.99346483580503e-06,
+      "loss": 0.8763,
+      "step": 120
+    },
+    {
+      "epoch": 0.05005294061026085,
+      "grad_norm": 0.15946023166179657,
+      "learning_rate": 9.989889390177467e-06,
+      "loss": 0.8688,
+      "step": 130
+    },
+    {
+      "epoch": 0.053903166811050146,
+      "grad_norm": 0.16416452825069427,
+      "learning_rate": 9.985537824862647e-06,
+      "loss": 0.8645,
+      "step": 140
+    },
+    {
+      "epoch": 0.057753393011839446,
+      "grad_norm": 0.18279142677783966,
+      "learning_rate": 9.980410816696317e-06,
+      "loss": 0.8609,
+      "step": 150
+    },
+    {
+      "epoch": 0.06160361921262874,
+      "grad_norm": 0.17308710515499115,
+      "learning_rate": 9.97450916312541e-06,
+      "loss": 0.8546,
+      "step": 160
+    },
+    {
+      "epoch": 0.06545384541341803,
+      "grad_norm": 0.1441011279821396,
+      "learning_rate": 9.967833782084004e-06,
+      "loss": 0.8523,
+      "step": 170
+    },
+    {
+      "epoch": 0.06930407161420733,
+      "grad_norm": 0.1650390326976776,
+      "learning_rate": 9.960385711850554e-06,
+      "loss": 0.8767,
+      "step": 180
+    },
+    {
+      "epoch": 0.07315429781499663,
+      "grad_norm": 0.17319875955581665,
+      "learning_rate": 9.952166110886394e-06,
+      "loss": 0.8764,
+      "step": 190
+    },
+    {
+      "epoch": 0.07700452401578593,
+      "grad_norm": 0.15899081528186798,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.8548,
+      "step": 200
+    },
+    {
+      "epoch": 0.08085475021657522,
+      "grad_norm": 0.1421322375535965,
+      "learning_rate": 9.933417550425952e-06,
+      "loss": 0.8607,
+      "step": 210
+    },
+    {
+      "epoch": 0.08470497641736452,
+      "grad_norm": 0.1572618931531906,
+      "learning_rate": 9.9228915070518e-06,
+      "loss": 0.8388,
+      "step": 220
+    },
+    {
+      "epoch": 0.08855520261815382,
+      "grad_norm": 0.17247164249420166,
+      "learning_rate": 9.911599764737632e-06,
+      "loss": 0.846,
+      "step": 230
+    },
+    {
+      "epoch": 0.09240542881894312,
+      "grad_norm": 0.18206538259983063,
+      "learning_rate": 9.899544079783606e-06,
+      "loss": 0.8325,
+      "step": 240
+    },
+    {
+      "epoch": 0.0962556550197324,
+      "grad_norm": 0.16972701251506805,
+      "learning_rate": 9.886726327312338e-06,
+      "loss": 0.8488,
+      "step": 250
+    },
+    {
+      "epoch": 0.1001058812205217,
+      "grad_norm": 0.14570847153663635,
+      "learning_rate": 9.87314850097724e-06,
+      "loss": 0.8522,
+      "step": 260
+    },
+    {
+      "epoch": 0.103956107421311,
+      "grad_norm": 0.15463249385356903,
+      "learning_rate": 9.858812712652452e-06,
+      "loss": 0.8372,
+      "step": 270
+    },
+    {
+      "epoch": 0.10780633362210029,
+      "grad_norm": 0.1552465260028839,
+      "learning_rate": 9.843721192104341e-06,
+      "loss": 0.8405,
+      "step": 280
+    },
+    {
+      "epoch": 0.11165655982288959,
+      "grad_norm": 0.15678738057613373,
+      "learning_rate": 9.827876286644698e-06,
+      "loss": 0.8532,
+      "step": 290
+    },
+    {
+      "epoch": 0.11550678602367889,
+      "grad_norm": 0.16120101511478424,
+      "learning_rate": 9.811280460765643e-06,
+      "loss": 0.8153,
+      "step": 300
+    },
+    {
+      "epoch": 0.11935701222446819,
+      "grad_norm": 0.1610136181116104,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.8317,
+      "step": 310
+    },
+    {
+      "epoch": 0.12320723842525748,
+      "grad_norm": 0.15977126359939575,
+      "learning_rate": 9.775846489301279e-06,
+      "loss": 0.8271,
+      "step": 320
+    },
+    {
+      "epoch": 0.12705746462604678,
+      "grad_norm": 0.1503450572490692,
+      "learning_rate": 9.757013855061157e-06,
+      "loss": 0.8298,
+      "step": 330
+    },
+    {
+      "epoch": 0.13090769082683607,
+      "grad_norm": 0.14118248224258423,
+      "learning_rate": 9.737441322234758e-06,
+      "loss": 0.8388,
+      "step": 340
+    },
+    {
+      "epoch": 0.13475791702762538,
+      "grad_norm": 0.16206443309783936,
+      "learning_rate": 9.717131935103605e-06,
+      "loss": 0.8436,
+      "step": 350
+    },
+    {
+      "epoch": 0.13860814322841467,
+      "grad_norm": 0.1374741494655609,
+      "learning_rate": 9.696088852558398e-06,
+      "loss": 0.8278,
+      "step": 360
+    },
+    {
+      "epoch": 0.14245836942920398,
+      "grad_norm": 0.15329225361347198,
+      "learning_rate": 9.674315347607686e-06,
+      "loss": 0.8525,
+      "step": 370
+    },
+    {
+      "epoch": 0.14630859562999327,
+      "grad_norm": 0.14636710286140442,
+      "learning_rate": 9.651814806868796e-06,
+      "loss": 0.8209,
+      "step": 380
+    },
+    {
+      "epoch": 0.15015882183078255,
+      "grad_norm": 0.14734572172164917,
+      "learning_rate": 9.62859073004107e-06,
+      "loss": 0.8265,
+      "step": 390
+    },
+    {
+      "epoch": 0.15400904803157187,
+      "grad_norm": 0.14635911583900452,
+      "learning_rate": 9.604646729361547e-06,
+      "loss": 0.8285,
+      "step": 400
+    },
+    {
+      "epoch": 0.15785927423236115,
+      "grad_norm": 0.13492073118686676,
+      "learning_rate": 9.579986529043099e-06,
+      "loss": 0.8084,
+      "step": 410
+    },
+    {
+      "epoch": 0.16170950043315044,
+      "grad_norm": 0.13152989745140076,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.8251,
+      "step": 420
+    },
+    {
+      "epoch": 0.16555972663393975,
+      "grad_norm": 0.15539216995239258,
+      "learning_rate": 9.528532982727281e-06,
+      "loss": 0.8171,
+      "step": 430
+    },
+    {
+      "epoch": 0.16940995283472904,
+      "grad_norm": 0.16300219297409058,
+      "learning_rate": 9.501747639735018e-06,
+      "loss": 0.8179,
+      "step": 440
+    },
+    {
+      "epoch": 0.17326017903551832,
+      "grad_norm": 0.1494835764169693,
+      "learning_rate": 9.474262101869272e-06,
+      "loss": 0.8016,
+      "step": 450
+    },
+    {
+      "epoch": 0.17711040523630764,
+      "grad_norm": 0.16662326455116272,
+      "learning_rate": 9.446080644188141e-06,
+      "loss": 0.8178,
+      "step": 460
+    },
+    {
+      "epoch": 0.18096063143709692,
+      "grad_norm": 0.15857912600040436,
+      "learning_rate": 9.417207649992017e-06,
+      "loss": 0.8231,
+      "step": 470
+    },
+    {
+      "epoch": 0.18481085763788624,
+      "grad_norm": 0.13895219564437866,
+      "learning_rate": 9.387647610141816e-06,
+      "loss": 0.817,
+      "step": 480
+    },
+    {
+      "epoch": 0.18866108383867553,
+      "grad_norm": 0.1402507871389389,
+      "learning_rate": 9.35740512236047e-06,
+      "loss": 0.8068,
+      "step": 490
+    },
+    {
+      "epoch": 0.1925113100394648,
+      "grad_norm": 0.13729862868785858,
+      "learning_rate": 9.326484890517809e-06,
+      "loss": 0.8087,
+      "step": 500
+    },
+    {
+      "epoch": 0.19636153624025413,
+      "grad_norm": 0.14094950258731842,
+      "learning_rate": 9.294891723898926e-06,
+      "loss": 0.7951,
+      "step": 510
+    },
+    {
+      "epoch": 0.2002117624410434,
+      "grad_norm": 0.1436399221420288,
+      "learning_rate": 9.262630536456153e-06,
+      "loss": 0.8362,
+      "step": 520
+    },
+    {
+      "epoch": 0.2040619886418327,
+      "grad_norm": 0.15702266991138458,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.8017,
+      "step": 530
+    },
+    {
+      "epoch": 0.207912214842622,
+      "grad_norm": 0.15012787282466888,
+      "learning_rate": 9.196124273642422e-06,
+      "loss": 0.822,
+      "step": 540
+    },
+    {
+      "epoch": 0.2117624410434113,
+      "grad_norm": 0.1375962346792221,
+      "learning_rate": 9.161889542552842e-06,
+      "loss": 0.8259,
+      "step": 550
+    },
+    {
+      "epoch": 0.21561266724420058,
+      "grad_norm": 0.1393718123435974,
+      "learning_rate": 9.127007477593193e-06,
+      "loss": 0.8265,
+      "step": 560
+    },
+    {
+      "epoch": 0.2194628934449899,
+      "grad_norm": 0.1519121676683426,
+      "learning_rate": 9.091483504265974e-06,
+      "loss": 0.8365,
+      "step": 570
+    },
+    {
+      "epoch": 0.22331311964577918,
+      "grad_norm": 0.14226572215557098,
+      "learning_rate": 9.055323147915116e-06,
+      "loss": 0.8114,
+      "step": 580
+    },
+    {
+      "epoch": 0.2271633458465685,
+      "grad_norm": 0.1538923680782318,
+      "learning_rate": 9.018532032866585e-06,
+      "loss": 0.8291,
+      "step": 590
+    },
+    {
+      "epoch": 0.23101357204735778,
+      "grad_norm": 0.16730016469955444,
+      "learning_rate": 8.981115881553579e-06,
+      "loss": 0.8236,
+      "step": 600
+    },
+    {
+      "epoch": 0.23486379824814707,
+      "grad_norm": 0.1477966606616974,
+      "learning_rate": 8.943080513626476e-06,
+      "loss": 0.8343,
+      "step": 610
+    },
+    {
+      "epoch": 0.23871402444893638,
+      "grad_norm": 0.1419568806886673,
+      "learning_rate": 8.904431845047641e-06,
+      "loss": 0.8266,
+      "step": 620
+    },
+    {
+      "epoch": 0.24256425064972567,
+      "grad_norm": 0.14538662135601044,
+      "learning_rate": 8.86517588717128e-06,
+      "loss": 0.8326,
+      "step": 630
+    },
+    {
+      "epoch": 0.24641447685051496,
+      "grad_norm": 0.132072314620018,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.8256,
+      "step": 640
+    },
+    {
+      "epoch": 0.25026470305130427,
+      "grad_norm": 0.13107925653457642,
+      "learning_rate": 8.784866620277314e-06,
+      "loss": 0.7815,
+      "step": 650
+    },
+    {
+      "epoch": 0.25411492925209356,
+      "grad_norm": 0.15237976610660553,
+      "learning_rate": 8.743825802439018e-06,
+      "loss": 0.8105,
+      "step": 660
+    },
+    {
+      "epoch": 0.25796515545288284,
+      "grad_norm": 0.18153919279575348,
+      "learning_rate": 8.702202675718956e-06,
+      "loss": 0.8245,
+      "step": 670
+    },
+    {
+      "epoch": 0.26181538165367213,
+      "grad_norm": 0.12988735735416412,
+      "learning_rate": 8.660003714113959e-06,
+      "loss": 0.8006,
+      "step": 680
+    },
+    {
+      "epoch": 0.26566560785446147,
+      "grad_norm": 0.1327754259109497,
+      "learning_rate": 8.617235481185323e-06,
+      "loss": 0.8142,
+      "step": 690
+    },
+    {
+      "epoch": 0.26951583405525076,
+      "grad_norm": 0.14275160431861877,
+      "learning_rate": 8.573904629037931e-06,
+      "loss": 0.8207,
+      "step": 700
+    },
+    {
+      "epoch": 0.27336606025604004,
+      "grad_norm": 0.1348872184753418,
+      "learning_rate": 8.530017897285585e-06,
+      "loss": 0.8054,
+      "step": 710
+    },
+    {
+      "epoch": 0.27721628645682933,
+      "grad_norm": 0.12930923700332642,
+      "learning_rate": 8.485582112002746e-06,
+      "loss": 0.8078,
+      "step": 720
+    },
+    {
+      "epoch": 0.2810665126576186,
+      "grad_norm": 0.1490335315465927,
+      "learning_rate": 8.440604184662811e-06,
+      "loss": 0.8183,
+      "step": 730
+    },
+    {
+      "epoch": 0.28491673885840796,
+      "grad_norm": 0.1500050127506256,
+      "learning_rate": 8.395091111063118e-06,
+      "loss": 0.7996,
+      "step": 740
+    },
+    {
+      "epoch": 0.28876696505919724,
+      "grad_norm": 0.1399359405040741,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.8007,
+      "step": 750
+    },
+    {
+      "epoch": 0.29261719125998653,
+      "grad_norm": 0.14406132698059082,
+      "learning_rate": 8.302487923351842e-06,
+      "loss": 0.7918,
+      "step": 760
+    },
+    {
+      "epoch": 0.2964674174607758,
+      "grad_norm": 0.15483282506465912,
+      "learning_rate": 8.255412212597021e-06,
+      "loss": 0.8109,
+      "step": 770
+    },
+    {
+      "epoch": 0.3003176436615651,
+      "grad_norm": 0.14815092086791992,
+      "learning_rate": 8.207830160055688e-06,
+      "loss": 0.8041,
+      "step": 780
+    },
+    {
+      "epoch": 0.3041678698623544,
+      "grad_norm": 0.1629728376865387,
+      "learning_rate": 8.159749166566786e-06,
+      "loss": 0.794,
+      "step": 790
+    },
+    {
+      "epoch": 0.30801809606314373,
+      "grad_norm": 0.15949465334415436,
+      "learning_rate": 8.11117671057377e-06,
+      "loss": 0.8199,
+      "step": 800
+    },
+    {
+      "epoch": 0.311868322263933,
+      "grad_norm": 0.14297303557395935,
+      "learning_rate": 8.062120346961404e-06,
+      "loss": 0.8069,
+      "step": 810
+    },
+    {
+      "epoch": 0.3157185484647223,
+      "grad_norm": 0.14033299684524536,
+      "learning_rate": 8.012587705880705e-06,
+      "loss": 0.8085,
+      "step": 820
+    },
+    {
+      "epoch": 0.3195687746655116,
+      "grad_norm": 0.14030049741268158,
+      "learning_rate": 7.962586491562146e-06,
+      "loss": 0.7961,
+      "step": 830
+    },
+    {
+      "epoch": 0.3234190008663009,
+      "grad_norm": 0.141440749168396,
+      "learning_rate": 7.912124481117357e-06,
+      "loss": 0.8013,
+      "step": 840
+    },
+    {
+      "epoch": 0.3272692270670902,
+      "grad_norm": 0.13020005822181702,
+      "learning_rate": 7.861209523329494e-06,
+      "loss": 0.8009,
+      "step": 850
+    },
+    {
+      "epoch": 0.3311194532678795,
+      "grad_norm": 0.1401718407869339,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.7946,
+      "step": 860
+    },
+    {
+      "epoch": 0.3349696794686688,
+      "grad_norm": 0.1414349526166916,
+      "learning_rate": 7.758052511879035e-06,
+      "loss": 0.7924,
+      "step": 870
+    },
+    {
+      "epoch": 0.3388199056694581,
+      "grad_norm": 0.12491948902606964,
+      "learning_rate": 7.705826503098642e-06,
+      "loss": 0.7988,
+      "step": 880
+    },
+    {
+      "epoch": 0.34267013187024736,
+      "grad_norm": 0.13912899792194366,
+      "learning_rate": 7.653179634243979e-06,
+      "loss": 0.79,
+      "step": 890
+    },
+    {
+      "epoch": 0.34652035807103665,
+      "grad_norm": 0.13653822243213654,
+      "learning_rate": 7.600120093927696e-06,
+      "loss": 0.7945,
+      "step": 900
+    },
+    {
+      "epoch": 0.350370584271826,
+      "grad_norm": 0.14164769649505615,
+      "learning_rate": 7.546656134948729e-06,
+      "loss": 0.795,
+      "step": 910
+    },
+    {
+      "epoch": 0.3542208104726153,
+      "grad_norm": 0.1413029432296753,
+      "learning_rate": 7.492796073008661e-06,
+      "loss": 0.783,
+      "step": 920
+    },
+    {
+      "epoch": 0.35807103667340456,
+      "grad_norm": 0.1440911591053009,
+      "learning_rate": 7.438548285418326e-06,
+      "loss": 0.7756,
+      "step": 930
+    },
+    {
+      "epoch": 0.36192126287419385,
+      "grad_norm": 0.14994767308235168,
+      "learning_rate": 7.3839212097948e-06,
+      "loss": 0.813,
+      "step": 940
+    },
+    {
+      "epoch": 0.36577148907498314,
+      "grad_norm": 0.13091829419136047,
+      "learning_rate": 7.328923342749039e-06,
+      "loss": 0.8238,
+      "step": 950
+    },
+    {
+      "epoch": 0.3696217152757725,
+      "grad_norm": 0.14799842238426208,
+      "learning_rate": 7.273563238564319e-06,
+      "loss": 0.8009,
+      "step": 960
+    },
+    {
+      "epoch": 0.37347194147656176,
+      "grad_norm": 0.14040197432041168,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.7826,
+      "step": 970
+    },
+    {
+      "epoch": 0.37732216767735105,
+      "grad_norm": 0.13302187621593475,
+      "learning_rate": 7.161790816280858e-06,
+      "loss": 0.7914,
+      "step": 980
+    },
+    {
+      "epoch": 0.38117239387814034,
+      "grad_norm": 0.1459832787513733,
+      "learning_rate": 7.1053958830920125e-06,
+      "loss": 0.7784,
+      "step": 990
+    },
+    {
+      "epoch": 0.3850226200789296,
+      "grad_norm": 0.13091851770877838,
+      "learning_rate": 7.0486734798799715e-06,
+      "loss": 0.8018,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3888728462797189,
+      "grad_norm": 0.14620435237884521,
+      "learning_rate": 6.991632429159708e-06,
+      "loss": 0.8211,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39272307248050825,
+      "grad_norm": 0.13758419454097748,
+      "learning_rate": 6.934281603008132e-06,
+      "loss": 0.7977,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39657329868129754,
+      "grad_norm": 0.13352185487747192,
+      "learning_rate": 6.876629921684143e-06,
+      "loss": 0.8087,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4004235248820868,
+      "grad_norm": 0.1390995979309082,
+      "learning_rate": 6.818686352241199e-06,
+      "loss": 0.8073,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4042737510828761,
+      "grad_norm": 0.1256643533706665,
+      "learning_rate": 6.760459907132573e-06,
+      "loss": 0.7979,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4081239772836654,
+      "grad_norm": 0.13719214498996735,
+      "learning_rate": 6.701959642809589e-06,
+      "loss": 0.8027,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41197420348445474,
+      "grad_norm": 0.13935986161231995,
+      "learning_rate": 6.643194658312987e-06,
+      "loss": 0.7855,
+      "step": 1070
+    },
+    {
+      "epoch": 0.415824429685244,
+      "grad_norm": 0.1446421593427658,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.8178,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4196746558860333,
+      "grad_norm": 0.1492662876844406,
+      "learning_rate": 6.5249071294110735e-06,
+      "loss": 0.7925,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4235248820868226,
+      "grad_norm": 0.15231294929981232,
+      "learning_rate": 6.465402983265274e-06,
+      "loss": 0.8136,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4273751082876119,
+      "grad_norm": 0.14420637488365173,
+      "learning_rate": 6.405670910603243e-06,
+      "loss": 0.8066,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43122533448840117,
+      "grad_norm": 0.15447750687599182,
+      "learning_rate": 6.345720202059291e-06,
+      "loss": 0.801,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4350755606891905,
+      "grad_norm": 0.14134666323661804,
+      "learning_rate": 6.285560182274008e-06,
+      "loss": 0.7881,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4389257868899798,
+      "grad_norm": 0.13190104067325592,
+      "learning_rate": 6.225200208443936e-06,
+      "loss": 0.7945,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4427760130907691,
+      "grad_norm": 0.12899965047836304,
+      "learning_rate": 6.16464966886616e-06,
+      "loss": 0.79,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44662623929155837,
+      "grad_norm": 0.15281686186790466,
+      "learning_rate": 6.103917981478065e-06,
+      "loss": 0.8179,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45047646549234766,
+      "grad_norm": 0.13590531051158905,
+      "learning_rate": 6.043014592392489e-06,
+      "loss": 0.8247,
+      "step": 1170
+    },
+    {
+      "epoch": 0.454326691693137,
+      "grad_norm": 0.12354939430952072,
+      "learning_rate": 5.981948974428485e-06,
+      "loss": 0.7964,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4581769178939263,
+      "grad_norm": 0.14219610393047333,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.7784,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46202714409471557,
+      "grad_norm": 0.14620943367481232,
+      "learning_rate": 5.8593690678282355e-06,
+      "loss": 0.7854,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46587737029550486,
+      "grad_norm": 0.13813823461532593,
+      "learning_rate": 5.797873845081301e-06,
+      "loss": 0.7914,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46972759649629414,
+      "grad_norm": 0.1335328370332718,
+      "learning_rate": 5.736254522269075e-06,
+      "loss": 0.8058,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47357782269708343,
+      "grad_norm": 0.13793028891086578,
+      "learning_rate": 5.6745206835658386e-06,
+      "loss": 0.808,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47742804889787277,
+      "grad_norm": 0.12669743597507477,
+      "learning_rate": 5.612681930957495e-06,
+      "loss": 0.7864,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48127827509866206,
+      "grad_norm": 0.13357707858085632,
+      "learning_rate": 5.550747882748095e-06,
+      "loss": 0.7841,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48512850129945134,
+      "grad_norm": 0.1415117383003235,
+      "learning_rate": 5.4887281720638195e-06,
+      "loss": 0.7834,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48897872750024063,
+      "grad_norm": 0.15428930521011353,
+      "learning_rate": 5.42663244535466e-06,
+      "loss": 0.775,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4928289537010299,
+      "grad_norm": 0.14762918651103973,
+      "learning_rate": 5.3644703608940205e-06,
+      "loss": 0.7808,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49667917990181926,
+      "grad_norm": 0.13139815628528595,
+      "learning_rate": 5.302251587276491e-06,
+      "loss": 0.8019,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5005294061026085,
+      "grad_norm": 0.14074364304542542,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.7943,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5043796323033978,
+      "grad_norm": 0.1478090137243271,
+      "learning_rate": 5.177682689530615e-06,
+      "loss": 0.8091,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5082298585041871,
+      "grad_norm": 0.1555299162864685,
+      "learning_rate": 5.11535194065619e-06,
+      "loss": 0.7851,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5120800847049765,
+      "grad_norm": 0.15312159061431885,
+      "learning_rate": 5.053003250119113e-06,
+      "loss": 0.7882,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5159303109057657,
+      "grad_norm": 0.13702252507209778,
+      "learning_rate": 4.9906463155383925e-06,
+      "loss": 0.793,
+      "step": 1340
+    },
+    {
+      "epoch": 0.519780537106555,
+      "grad_norm": 0.14805445075035095,
+      "learning_rate": 4.9282908358153046e-06,
+      "loss": 0.7822,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5236307633073443,
+      "grad_norm": 0.15615208446979523,
+      "learning_rate": 4.8659465096248345e-06,
+      "loss": 0.7974,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5274809895081336,
+      "grad_norm": 0.1494620442390442,
+      "learning_rate": 4.803623033907169e-06,
+      "loss": 0.8038,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5313312157089229,
+      "grad_norm": 0.12665468454360962,
+      "learning_rate": 4.741330102359441e-06,
+      "loss": 0.8062,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5351814419097122,
+      "grad_norm": 0.14424964785575867,
+      "learning_rate": 4.679077403927996e-06,
+      "loss": 0.7819,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5390316681105015,
+      "grad_norm": 0.1356794834136963,
+      "learning_rate": 4.616874621301373e-06,
+      "loss": 0.7886,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5428818943112907,
+      "grad_norm": 0.14265860617160797,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.8039,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5467321205120801,
+      "grad_norm": 0.15258218348026276,
+      "learning_rate": 4.492657493892822e-06,
+      "loss": 0.7897,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5505823467128694,
+      "grad_norm": 0.14812804758548737,
+      "learning_rate": 4.430662469650999e-06,
+      "loss": 0.7993,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5544325729136587,
+      "grad_norm": 0.14322498440742493,
+      "learning_rate": 4.368755999289118e-06,
+      "loss": 0.8012,
+      "step": 1440
+    },
+    {
+      "epoch": 0.558282799114448,
+      "grad_norm": 0.173824742436409,
+      "learning_rate": 4.306947711643949e-06,
+      "loss": 0.7859,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5621330253152372,
+      "grad_norm": 0.12926770746707916,
+      "learning_rate": 4.245247220281068e-06,
+      "loss": 0.7898,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5659832515160266,
+      "grad_norm": 0.13061374425888062,
+      "learning_rate": 4.183664121999587e-06,
+      "loss": 0.786,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5698334777168159,
+      "grad_norm": 0.14013013243675232,
+      "learning_rate": 4.122207995339474e-06,
+      "loss": 0.7857,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5736837039176051,
+      "grad_norm": 0.13764025270938873,
+      "learning_rate": 4.060888399091734e-06,
+      "loss": 0.7908,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5775339301183945,
+      "grad_norm": 0.13493385910987854,
+      "learning_rate": 3.999714870811639e-06,
+      "loss": 0.7982,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5813841563191837,
+      "grad_norm": 0.13432228565216064,
+      "learning_rate": 3.938696925335275e-06,
+      "loss": 0.7879,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5852343825199731,
+      "grad_norm": 0.14636914432048798,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.797,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5890846087207623,
+      "grad_norm": 0.15518972277641296,
+      "learning_rate": 3.817165719666382e-06,
+      "loss": 0.7983,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5929348349215516,
+      "grad_norm": 0.13408838212490082,
+      "learning_rate": 3.756671362249841e-06,
+      "loss": 0.8032,
+      "step": 1540
+    },
+    {
+      "epoch": 0.596785061122341,
+      "grad_norm": 0.13932813704013824,
+      "learning_rate": 3.6963703902488907e-06,
+      "loss": 0.797,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6006352873231302,
+      "grad_norm": 0.13751046359539032,
+      "learning_rate": 3.6362721827835655e-06,
+      "loss": 0.8119,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6044855135239195,
+      "grad_norm": 0.15106604993343353,
+      "learning_rate": 3.576386087436212e-06,
+      "loss": 0.7967,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6083357397247088,
+      "grad_norm": 0.1363649070262909,
+      "learning_rate": 3.5167214187975913e-06,
+      "loss": 0.7981,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6121859659254981,
+      "grad_norm": 0.12219467014074326,
+      "learning_rate": 3.457287457018086e-06,
+      "loss": 0.7962,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6160361921262875,
+      "grad_norm": 0.13305144011974335,
+      "learning_rate": 3.3980934463643003e-06,
+      "loss": 0.779,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6198864183270767,
+      "grad_norm": 0.12293121963739395,
+      "learning_rate": 3.3391485937812053e-06,
+      "loss": 0.7628,
+      "step": 1610
+    },
+    {
+      "epoch": 0.623736644527866,
+      "grad_norm": 0.1354309320449829,
+      "learning_rate": 3.280462067460115e-06,
+      "loss": 0.7651,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6275868707286553,
+      "grad_norm": 0.13088493049144745,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.7999,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6314370969294446,
+      "grad_norm": 0.12868967652320862,
+      "learning_rate": 3.163900464051089e-06,
+      "loss": 0.7828,
+      "step": 1640
+    },
+    {
+      "epoch": 0.635287323130234,
+      "grad_norm": 0.14205335080623627,
+      "learning_rate": 3.1060435167748826e-06,
+      "loss": 0.7926,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6391375493310232,
+      "grad_norm": 0.13787361979484558,
+      "learning_rate": 3.048481152564253e-06,
+      "loss": 0.7789,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6429877755318125,
+      "grad_norm": 0.1426582634449005,
+      "learning_rate": 2.9912223245804017e-06,
+      "loss": 0.8006,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6468380017326018,
+      "grad_norm": 0.13891686499118805,
+      "learning_rate": 2.934275938772981e-06,
+      "loss": 0.7888,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6506882279333911,
+      "grad_norm": 0.1449626088142395,
+      "learning_rate": 2.8776508524948676e-06,
+      "loss": 0.7763,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6545384541341804,
+      "grad_norm": 0.12573854625225067,
+      "learning_rate": 2.821355873124503e-06,
+      "loss": 0.7747,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6583886803349697,
+      "grad_norm": 0.14211127161979675,
+      "learning_rate": 2.7653997566960144e-06,
+      "loss": 0.7934,
+      "step": 1710
+    },
+    {
+      "epoch": 0.662238906535759,
+      "grad_norm": 0.1389818638563156,
+      "learning_rate": 2.709791206537309e-06,
+      "loss": 0.7839,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6660891327365482,
+      "grad_norm": 0.12983375787734985,
+      "learning_rate": 2.6545388719163747e-06,
+      "loss": 0.7897,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6699393589373376,
+      "grad_norm": 0.14210471510887146,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.7935,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6737895851381268,
+      "grad_norm": 0.1384427696466446,
+      "learning_rate": 2.545137167997007e-06,
+      "loss": 0.7878,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6776398113389162,
+      "grad_norm": 0.12455567717552185,
+      "learning_rate": 2.4910048148705972e-06,
+      "loss": 0.7744,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6814900375397055,
+      "grad_norm": 0.13967594504356384,
+      "learning_rate": 2.4372627069793327e-06,
+      "loss": 0.8141,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6853402637404947,
+      "grad_norm": 0.14710889756679535,
+      "learning_rate": 2.3839192032876486e-06,
+      "loss": 0.7984,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6891904899412841,
+      "grad_norm": 0.1304047554731369,
+      "learning_rate": 2.3309826007617047e-06,
+      "loss": 0.8105,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6930407161420733,
+      "grad_norm": 0.13599811494350433,
+      "learning_rate": 2.278461133078881e-06,
+      "loss": 0.777,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6968909423428626,
+      "grad_norm": 0.13404828310012817,
+      "learning_rate": 2.2263629693471258e-06,
+      "loss": 0.7706,
+      "step": 1810
+    },
+    {
+      "epoch": 0.700741168543652,
+      "grad_norm": 0.13723786175251007,
+      "learning_rate": 2.1746962128343414e-06,
+      "loss": 0.7957,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7045913947444412,
+      "grad_norm": 0.1350916475057602,
+      "learning_rate": 2.1234688997080183e-06,
+      "loss": 0.7896,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7084416209452306,
+      "grad_norm": 0.1407240480184555,
+      "learning_rate": 2.0726889977853053e-06,
+      "loss": 0.7881,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7122918471460198,
+      "grad_norm": 0.13579164445400238,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.7974,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7161420733468091,
+      "grad_norm": 0.13240160048007965,
+      "learning_rate": 1.9725029496425814e-06,
+      "loss": 0.7936,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7199922995475985,
+      "grad_norm": 0.13771021366119385,
+      "learning_rate": 1.923112386205727e-06,
+      "loss": 0.7897,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7238425257483877,
+      "grad_norm": 0.1487494707107544,
+      "learning_rate": 1.874200397115079e-06,
+      "loss": 0.7937,
+      "step": 1880
+    },
+    {
+      "epoch": 0.727692751949177,
+      "grad_norm": 0.1375429332256317,
+      "learning_rate": 1.825774590065857e-06,
+      "loss": 0.7813,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7315429781499663,
+      "grad_norm": 0.1419457644224167,
+      "learning_rate": 1.7778424971332813e-06,
+      "loss": 0.7893,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7353932043507556,
+      "grad_norm": 0.12965650856494904,
+      "learning_rate": 1.7304115736010407e-06,
+      "loss": 0.7906,
+      "step": 1910
+    },
+    {
+      "epoch": 0.739243430551545,
+      "grad_norm": 0.14045630395412445,
+      "learning_rate": 1.6834891968017108e-06,
+      "loss": 0.7805,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7430936567523342,
+      "grad_norm": 0.1372210681438446,
+      "learning_rate": 1.6370826649692884e-06,
+      "loss": 0.7977,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7469438829531235,
+      "grad_norm": 0.13613587617874146,
+      "learning_rate": 1.5911991961040463e-06,
+      "loss": 0.8067,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7507941091539128,
+      "grad_norm": 0.14294028282165527,
+      "learning_rate": 1.545845926849851e-06,
+      "loss": 0.7933,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7546443353547021,
+      "grad_norm": 0.14036765694618225,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.8092,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7584945615554913,
+      "grad_norm": 0.1360415518283844,
+      "learning_rate": 1.4567581203207238e-06,
+      "loss": 0.802,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7623447877562807,
+      "grad_norm": 0.13754723966121674,
+      "learning_rate": 1.413037439625593e-06,
+      "loss": 0.7896,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76619501395707,
+      "grad_norm": 0.15682174265384674,
+      "learning_rate": 1.3698746695458847e-06,
+      "loss": 0.7914,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7700452401578592,
+      "grad_norm": 0.13540446758270264,
+      "learning_rate": 1.3272765235521745e-06,
+      "loss": 0.7899,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7738954663586486,
+      "grad_norm": 0.1260048747062683,
+      "learning_rate": 1.2852496272942877e-06,
+      "loss": 0.7882,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7777456925594378,
+      "grad_norm": 0.13515381515026093,
+      "learning_rate": 1.2438005175707396e-06,
+      "loss": 0.7766,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7815959187602272,
+      "grad_norm": 0.1370023936033249,
+      "learning_rate": 1.2029356413120297e-06,
+      "loss": 0.7655,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7854461449610165,
+      "grad_norm": 0.14078304171562195,
+      "learning_rate": 1.1626613545778786e-06,
+      "loss": 0.7822,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7892963711618057,
+      "grad_norm": 0.12863022089004517,
+      "learning_rate": 1.1229839215686294e-06,
+      "loss": 0.7875,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7931465973625951,
+      "grad_norm": 0.12404927611351013,
+      "learning_rate": 1.0839095136509188e-06,
+      "loss": 0.7714,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7969968235633843,
+      "grad_norm": 0.15145590901374817,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.7852,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8008470497641736,
+      "grad_norm": 0.1352020502090454,
+      "learning_rate": 1.007593988643402e-06,
+      "loss": 0.7672,
+      "step": 2080
+    },
+    {
+      "epoch": 0.804697275964963,
+      "grad_norm": 0.1346360146999359,
+      "learning_rate": 9.703647415524613e-07,
+      "loss": 0.7926,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8085475021657522,
+      "grad_norm": 0.13695302605628967,
+      "learning_rate": 9.337622577045485e-07,
+      "loss": 0.786,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8123977283665416,
+      "grad_norm": 0.132505863904953,
+      "learning_rate": 8.977922301934588e-07,
+      "loss": 0.7963,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8162479545673308,
+      "grad_norm": 0.13822650909423828,
+      "learning_rate": 8.624602537417009e-07,
+      "loss": 0.7949,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8200981807681201,
+      "grad_norm": 0.13586057722568512,
+      "learning_rate": 8.277718238303123e-07,
+      "loss": 0.7817,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8239484069689095,
+      "grad_norm": 0.15184904634952545,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.7927,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8277986331696987,
+      "grad_norm": 0.144210085272789,
+      "learning_rate": 7.603470842324223e-07,
+      "loss": 0.7785,
+      "step": 2150
+    },
+    {
+      "epoch": 0.831648859370488,
+      "grad_norm": 0.13371028006076813,
+      "learning_rate": 7.276212616857575e-07,
+      "loss": 0.7925,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8354990855712773,
+      "grad_norm": 0.1298682987689972,
+      "learning_rate": 6.955599583279837e-07,
+      "loss": 0.7966,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8393493117720666,
+      "grad_norm": 0.13028740882873535,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.7842,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8431995379728559,
+      "grad_norm": 0.14026489853858948,
+      "learning_rate": 6.334507521075711e-07,
+      "loss": 0.7964,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8470497641736452,
+      "grad_norm": 0.14467112720012665,
+      "learning_rate": 6.034125096149179e-07,
+      "loss": 0.7787,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8508999903744345,
+      "grad_norm": 0.14305898547172546,
+      "learning_rate": 5.740581055485722e-07,
+      "loss": 0.7668,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8547502165752238,
+      "grad_norm": 0.1252230405807495,
+      "learning_rate": 5.453921056471961e-07,
+      "loss": 0.791,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8586004427760131,
+      "grad_norm": 0.14678342640399933,
+      "learning_rate": 5.174189685761288e-07,
+      "loss": 0.7811,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8624506689768023,
+      "grad_norm": 0.13936232030391693,
+      "learning_rate": 4.901430452338979e-07,
+      "loss": 0.8002,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8663008951775917,
+      "grad_norm": 0.13745880126953125,
+      "learning_rate": 4.6356857807548286e-07,
+      "loss": 0.7831,
+      "step": 2250
+    },
+    {
+      "epoch": 0.870151121378381,
+      "grad_norm": 0.14044199883937836,
+      "learning_rate": 4.376997004524541e-07,
+      "loss": 0.7722,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8740013475791703,
+      "grad_norm": 0.13185228407382965,
+      "learning_rate": 4.12540435970073e-07,
+      "loss": 0.7787,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8778515737799596,
+      "grad_norm": 0.12850144505500793,
+      "learning_rate": 3.880946978614703e-07,
+      "loss": 0.7908,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8817017999807488,
+      "grad_norm": 0.1313028335571289,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.7964,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8855520261815382,
+      "grad_norm": 0.12733328342437744,
+      "learning_rate": 3.41358898202781e-07,
+      "loss": 0.7762,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8894022523823275,
+      "grad_norm": 0.12479226291179657,
+      "learning_rate": 3.190761058667741e-07,
+      "loss": 0.7786,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8932524785831167,
+      "grad_norm": 0.12567076086997986,
+      "learning_rate": 2.9752137720206463e-07,
+      "loss": 0.7893,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8971027047839061,
+      "grad_norm": 0.13714465498924255,
+      "learning_rate": 2.7669806479785067e-07,
+      "loss": 0.7994,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9009529309846953,
+      "grad_norm": 0.13255982100963593,
+      "learning_rate": 2.566094074799763e-07,
+      "loss": 0.7911,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9048031571854847,
+      "grad_norm": 0.1359822303056717,
+      "learning_rate": 2.3725852980716557e-07,
+      "loss": 0.7709,
+      "step": 2350
+    },
+    {
+      "epoch": 0.908653383386274,
+      "grad_norm": 0.12995357811450958,
+      "learning_rate": 2.1864844158503983e-07,
+      "loss": 0.7901,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9125036095870632,
+      "grad_norm": 0.14315855503082275,
+      "learning_rate": 2.007820373979702e-07,
+      "loss": 0.7739,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9163538357878526,
+      "grad_norm": 0.14017386734485626,
+      "learning_rate": 1.8366209615886532e-07,
+      "loss": 0.7797,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9202040619886418,
+      "grad_norm": 0.12661603093147278,
+      "learning_rate": 1.6729128067693613e-07,
+      "loss": 0.7977,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9240542881894311,
+      "grad_norm": 0.13592569530010223,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.788,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9279045143902205,
+      "grad_norm": 0.1343042254447937,
+      "learning_rate": 1.3680709523610058e-07,
+      "loss": 0.7768,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9317547405910097,
+      "grad_norm": 0.12327102571725845,
+      "learning_rate": 1.2269846674030418e-07,
+      "loss": 0.8008,
+      "step": 2420
+    },
+    {
+      "epoch": 0.935604966791799,
+      "grad_norm": 0.13680343329906464,
+      "learning_rate": 1.0934844619042595e-07,
+      "loss": 0.7965,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9394551929925883,
+      "grad_norm": 0.13609623908996582,
+      "learning_rate": 9.675911002803961e-08,
+      "loss": 0.7849,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9433054191933776,
+      "grad_norm": 0.1429198980331421,
+      "learning_rate": 8.493241637904381e-08,
+      "loss": 0.7844,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9471556453941669,
+      "grad_norm": 0.140077605843544,
+      "learning_rate": 7.387020474909689e-08,
+      "loss": 0.8025,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9510058715949562,
+      "grad_norm": 0.13803914189338684,
+      "learning_rate": 6.357419573750856e-08,
+      "loss": 0.7886,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9548560977957455,
+      "grad_norm": 0.13122986257076263,
+      "learning_rate": 5.4045990769613365e-08,
+      "loss": 0.784,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9587063239965348,
+      "grad_norm": 0.14976899325847626,
+      "learning_rate": 4.5287071847694434e-08,
+      "loss": 0.7864,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9625565501973241,
+      "grad_norm": 0.1372964084148407,
+      "learning_rate": 3.7298801320467817e-08,
+      "loss": 0.7768,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9664067763981133,
+      "grad_norm": 0.1306471824645996,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.7756,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9702570025989027,
+      "grad_norm": 0.12621468305587769,
+      "learning_rate": 2.3639055324410355e-08,
+      "loss": 0.7905,
+      "step": 2520
+    },
+    {
+      "epoch": 0.974107228799692,
+      "grad_norm": 0.13574396073818207,
+      "learning_rate": 1.7969704471369477e-08,
+      "loss": 0.7789,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9779574550004813,
+      "grad_norm": 0.127005472779274,
+      "learning_rate": 1.3075250914142413e-08,
+      "loss": 0.7841,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9818076812012706,
+      "grad_norm": 0.12407071888446808,
+      "learning_rate": 8.956455928476049e-09,
+      "loss": 0.7947,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9856579074020598,
+      "grad_norm": 0.12473104149103165,
+      "learning_rate": 5.6139601453786275e-09,
+      "loss": 0.7746,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9895081336028492,
+      "grad_norm": 0.13957831263542175,
+      "learning_rate": 3.0482834514805602e-09,
+      "loss": 0.7774,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9933583598036385,
+      "grad_norm": 0.12257251888513565,
+      "learning_rate": 1.2598249081713366e-09,
+      "loss": 0.7671,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9972085860044277,
+      "grad_norm": 0.1390281319618225,
+      "learning_rate": 2.488626895302826e-10,
+      "loss": 0.7932,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9999037443449803,
+      "step": 2597,
+      "total_flos": 4.083574038318809e+18,
+      "train_loss": 0.8208777617159649,
+      "train_runtime": 16092.3344,
+      "train_samples_per_second": 41.318,
+      "train_steps_per_second": 0.161
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2597,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.083574038318809e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3/mc2_h14_378/finetune/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c9264cb3b313ce2f7f9159ba4fdb57360a153f5abaffe91f17c726b4c78f91d
+size 6136

continuous/qwen3/mc2_h14_378/pretrain/config.json

@@ -0,0 +1,60 @@
+{
+  "_attn_implementation_autoset": true,
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "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,
+  "max_position_embeddings": 40960,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1280,
+  "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-H-14-worldwide-378",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "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",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "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_h14_378.pt",
+  "vocab_size": 151936
+}

continuous/qwen3/mc2_h14_378/pretrain/mm_projector.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6783f3e38fa38d7c6de43808e93b4a7f53f9bc89d4d87fff8283aabafa96de97
+size 27281464

continuous/qwen3/mc2_h14_378/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": 0.42371100187301636,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 9.1504,
+      "step": 10
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 0.1479865461587906,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 5.6256,
+      "step": 20
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.04121469706296921,
+      "learning_rate": 0.0009999907688176173,
+      "loss": 4.9642,
+      "step": 30
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.05725841224193573,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 4.8181,
+      "step": 40
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.02850869670510292,
+      "learning_rate": 0.0009959345572940772,
+      "loss": 4.7523,
+      "step": 50
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.015091827139258385,
+      "learning_rate": 0.0009911550081155983,
+      "loss": 4.7199,
+      "step": 60
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.07685502618551254,
+      "learning_rate": 0.0009845624347566433,
+      "loss": 4.6826,
+      "step": 70
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.018897874280810356,
+      "learning_rate": 0.0009761811727013547,
+      "loss": 4.6677,
+      "step": 80
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.03871753811836243,
+      "learning_rate": 0.0009660421601067666,
+      "loss": 4.6463,
+      "step": 90
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.04579046368598938,
+      "learning_rate": 0.0009541828235992926,
+      "loss": 4.6295,
+      "step": 100
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.05107562243938446,
+      "learning_rate": 0.0009406469401200151,
+      "loss": 4.5952,
+      "step": 110
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.0480436310172081,
+      "learning_rate": 0.0009254844753287492,
+      "loss": 4.5763,
+      "step": 120
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.03307349979877472,
+      "learning_rate": 0.0009087513991633924,
+      "loss": 4.5473,
+      "step": 130
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.04237901791930199,
+      "learning_rate": 0.0008905094792353917,
+      "loss": 4.5216,
+      "step": 140
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.04316609725356102,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 4.509,
+      "step": 150
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.07799968123435974,
+      "learning_rate": 0.0008497737783108237,
+      "loss": 4.4855,
+      "step": 160
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.04152051731944084,
+      "learning_rate": 0.0008274303669726426,
+      "loss": 4.4659,
+      "step": 170
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.06449169665575027,
+      "learning_rate": 0.0008038782961218137,
+      "loss": 4.4396,
+      "step": 180
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.04165741801261902,
+      "learning_rate": 0.000779204504653894,
+      "loss": 4.4288,
+      "step": 190
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.04213886335492134,
+      "learning_rate": 0.0007535000721258791,
+      "loss": 4.3931,
+      "step": 200
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.038481734693050385,
+      "learning_rate": 0.00072685988254984,
+      "loss": 4.3732,
+      "step": 210
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.05347040295600891,
+      "learning_rate": 0.0006993822741429906,
+      "loss": 4.3779,
+      "step": 220
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.04301167652010918,
+      "learning_rate": 0.0006711686763270818,
+      "loss": 4.3656,
+      "step": 230
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.043588053435087204,
+      "learning_rate": 0.0006423232353170831,
+      "loss": 4.3415,
+      "step": 240
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.04874352738261223,
+      "learning_rate": 0.0006129524296812335,
+      "loss": 4.3171,
+      "step": 250
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.03259924054145813,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 4.3232,
+      "step": 260
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.034871794283390045,
+      "learning_rate": 0.0005530699351157809,
+      "loss": 4.2963,
+      "step": 270
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.03127312287688255,
+      "learning_rate": 0.0005227792933277943,
+      "loss": 4.2866,
+      "step": 280
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.03340888395905495,
+      "learning_rate": 0.0004924045652352092,
+      "loss": 4.2688,
+      "step": 290
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.03931131958961487,
+      "learning_rate": 0.0004620578745374633,
+      "loss": 4.2634,
+      "step": 300
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.03873540833592415,
+      "learning_rate": 0.0004318512414382059,
+      "loss": 4.2744,
+      "step": 310
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.029258327558636665,
+      "learning_rate": 0.0004018961691397148,
+      "loss": 4.2621,
+      "step": 320
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.026353415101766586,
+      "learning_rate": 0.0003723032322457458,
+      "loss": 4.2395,
+      "step": 330
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.029283620417118073,
+      "learning_rate": 0.0003431816685921625,
+      "loss": 4.2398,
+      "step": 340
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.02842608094215393,
+      "learning_rate": 0.0003146389760120416,
+      "loss": 4.2385,
+      "step": 350
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.03155658766627312,
+      "learning_rate": 0.00028678051552373485,
+      "loss": 4.2442,
+      "step": 360
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.024778639897704124,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 4.223,
+      "step": 370
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.030074341222643852,
+      "learning_rate": 0.0002335247266004629,
+      "loss": 4.2374,
+      "step": 380
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.03499709069728851,
+      "learning_rate": 0.000208323983828828,
+      "loss": 4.2168,
+      "step": 390
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.030278515070676804,
+      "learning_rate": 0.0001841999188095224,
+      "loss": 4.2196,
+      "step": 400
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.03196336328983307,
+      "learning_rate": 0.00016124158186777676,
+      "loss": 4.217,
+      "step": 410
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.027287527918815613,
+      "learning_rate": 0.0001395337202205211,
+      "loss": 4.2089,
+      "step": 420
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.024377185851335526,
+      "learning_rate": 0.00011915646514489292,
+      "loss": 4.2125,
+      "step": 430
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.025281816720962524,
+      "learning_rate": 0.00010018503618578818,
+      "loss": 4.2202,
+      "step": 440
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.024551551789045334,
+      "learning_rate": 8.268946349432583e-05,
+      "loss": 4.2089,
+      "step": 450
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.022201765328645706,
+      "learning_rate": 6.673432932217117e-05,
+      "loss": 4.2016,
+      "step": 460
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.02687288634479046,
+      "learning_rate": 5.2378529625954684e-05,
+      "loss": 4.1906,
+      "step": 470
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.02313816174864769,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 4.201,
+      "step": 480
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.02198879048228264,
+      "learning_rate": 2.8670803372379538e-05,
+      "loss": 4.1995,
+      "step": 490
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.0225325096398592,
+      "learning_rate": 1.9406390288876586e-05,
+      "loss": 4.2135,
+      "step": 500
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.022999776527285576,
+      "learning_rate": 1.191601558631461e-05,
+      "loss": 4.2037,
+      "step": 510
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.02260691672563553,
+      "learning_rate": 6.2273288462559305e-06,
+      "loss": 4.1925,
+      "step": 520
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.022257601842284203,
+      "learning_rate": 2.361328992552314e-06,
+      "loss": 4.1937,
+      "step": 530
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.021725045517086983,
+      "learning_rate": 3.322867770044202e-07,
+      "loss": 4.2009,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "step": 545,
+      "total_flos": 3.5902605350037094e+17,
+      "train_loss": 4.480776645065448,
+      "train_runtime": 6568.7705,
+      "train_samples_per_second": 84.967,
+      "train_steps_per_second": 0.083
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 545,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.5902605350037094e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3/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/qwen3/mc2_l14_224/finetune/config.json

@@ -0,0 +1,59 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "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,
+  "max_position_embeddings": 40960,
+  "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,
+  "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",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "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/qwen3/mc2_l14_224/finetune/generation_config.json

@@ -0,0 +1,13 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.6,
+  "top_k": 20,
+  "top_p": 0.95,
+  "transformers_version": "4.51.3"
+}

continuous/qwen3/mc2_l14_224/finetune/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b4fc0c4d118a3a5407320b682662ef78df7f2566e070df92d6712f674f88184
+size 4696673872

continuous/qwen3/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": "<|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/mc2_l14_224/finetune/tokenizer.json

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

continuous/qwen3/mc2_l14_224/finetune/tokenizer_config.json

@@ -0,0 +1,241 @@
+{
+  "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,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0].role == 'system' %}\n        {{- messages[0].content + '\\n\\n' }}\n    {%- endif %}\n    {{- \"# 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>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\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\" }}\n{%- else %}\n    {%- if messages[0].role == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0].content + '<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n    {%- set index = (messages|length - 1) - loop.index0 %}\n    {%- if ns.multi_step_tool and message.role == \"user\" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n        {%- set ns.multi_step_tool = false %}\n        {%- set ns.last_query_index = index %}\n    {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n    {%- if message.content is string %}\n        {%- set content = message.content %}\n    {%- else %}\n        {%- set content = '' %}\n    {%- endif %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set reasoning_content = '' %}\n        {%- if message.reasoning_content is string %}\n            {%- set reasoning_content = message.reasoning_content %}\n        {%- else %}\n            {%- if '</think>' in content %}\n                {%- set reasoning_content = content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n                {%- set content = content.split('</think>')[-1].lstrip('\\n') %}\n            {%- endif %}\n        {%- endif %}\n        {%- if loop.index0 > ns.last_query_index %}\n            {%- if loop.last or (not loop.last and reasoning_content) %}\n                {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n            {%- else %}\n                {{- '<|im_start|>' + message.role + '\\n' + content }}\n            {%- endif %}\n        {%- else %}\n            {{- '<|im_start|>' + message.role + '\\n' + content }}\n        {%- endif %}\n        {%- if message.tool_calls %}\n            {%- for tool_call in message.tool_calls %}\n                {%- if (loop.first and content) or (not loop.first) %}\n                    {{- '\\n' }}\n                {%- endif %}\n                {%- if tool_call.function %}\n                    {%- set tool_call = tool_call.function %}\n                {%- endif %}\n                {{- '<tool_call>\\n{\"name\": \"' }}\n                {{- tool_call.name }}\n                {{- '\", \"arguments\": ' }}\n                {%- if tool_call.arguments is string %}\n                    {{- tool_call.arguments }}\n                {%- else %}\n                    {{- tool_call.arguments | tojson }}\n                {%- endif %}\n                {{- '}\\n</tool_call>' }}\n            {%- endfor %}\n        {%- endif %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n    {%- if enable_thinking is defined and enable_thinking is false %}\n        {{- '<think>\\n\\n</think>\\n\\n' }}\n    {%- endif %}\n{%- endif %}",
+  "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/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": 19.79184341430664,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 2.6348,
+      "step": 10
+    },
+    {
+      "epoch": 0.007700452401578592,
+      "grad_norm": 6.341690540313721,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 2.0129,
+      "step": 20
+    },
+    {
+      "epoch": 0.01155067860236789,
+      "grad_norm": 4.526823997497559,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.51,
+      "step": 30
+    },
+    {
+      "epoch": 0.015400904803157185,
+      "grad_norm": 1.1280349493026733,
+      "learning_rate": 5e-06,
+      "loss": 1.1747,
+      "step": 40
+    },
+    {
+      "epoch": 0.019251131003946483,
+      "grad_norm": 0.7391446232795715,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.0718,
+      "step": 50
+    },
+    {
+      "epoch": 0.02310135720473578,
+      "grad_norm": 0.39864447712898254,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.986,
+      "step": 60
+    },
+    {
+      "epoch": 0.026951583405525073,
+      "grad_norm": 0.3084148168563843,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.9356,
+      "step": 70
+    },
+    {
+      "epoch": 0.03080180960631437,
+      "grad_norm": 0.2990218997001648,
+      "learning_rate": 9.999996111488723e-06,
+      "loss": 0.9309,
+      "step": 80
+    },
+    {
+      "epoch": 0.034652035807103666,
+      "grad_norm": 0.2780090272426605,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 0.8979,
+      "step": 90
+    },
+    {
+      "epoch": 0.038502262007892966,
+      "grad_norm": 0.24935807287693024,
+      "learning_rate": 9.998285264324193e-06,
+      "loss": 0.8789,
+      "step": 100
+    },
+    {
+      "epoch": 0.04235248820868226,
+      "grad_norm": 0.26973283290863037,
+      "learning_rate": 9.996263605626052e-06,
+      "loss": 0.86,
+      "step": 110
+    },
+    {
+      "epoch": 0.04620271440947156,
+      "grad_norm": 0.24616661667823792,
+      "learning_rate": 9.99346483580503e-06,
+      "loss": 0.8691,
+      "step": 120
+    },
+    {
+      "epoch": 0.05005294061026085,
+      "grad_norm": 0.25066259503364563,
+      "learning_rate": 9.989889390177467e-06,
+      "loss": 0.8648,
+      "step": 130
+    },
+    {
+      "epoch": 0.053903166811050146,
+      "grad_norm": 0.28245049715042114,
+      "learning_rate": 9.985537824862647e-06,
+      "loss": 0.8587,
+      "step": 140
+    },
+    {
+      "epoch": 0.057753393011839446,
+      "grad_norm": 0.29467615485191345,
+      "learning_rate": 9.980410816696317e-06,
+      "loss": 0.8449,
+      "step": 150
+    },
+    {
+      "epoch": 0.06160361921262874,
+      "grad_norm": 0.257530152797699,
+      "learning_rate": 9.97450916312541e-06,
+      "loss": 0.8491,
+      "step": 160
+    },
+    {
+      "epoch": 0.06545384541341803,
+      "grad_norm": 0.22230692207813263,
+      "learning_rate": 9.967833782084004e-06,
+      "loss": 0.85,
+      "step": 170
+    },
+    {
+      "epoch": 0.06930407161420733,
+      "grad_norm": 0.2579823136329651,
+      "learning_rate": 9.960385711850554e-06,
+      "loss": 0.877,
+      "step": 180
+    },
+    {
+      "epoch": 0.07315429781499663,
+      "grad_norm": 0.256746381521225,
+      "learning_rate": 9.952166110886394e-06,
+      "loss": 0.8668,
+      "step": 190
+    },
+    {
+      "epoch": 0.07700452401578593,
+      "grad_norm": 0.2952767312526703,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.8563,
+      "step": 200
+    },
+    {
+      "epoch": 0.08085475021657522,
+      "grad_norm": 0.24737772345542908,
+      "learning_rate": 9.933417550425952e-06,
+      "loss": 0.8509,
+      "step": 210
+    },
+    {
+      "epoch": 0.08470497641736452,
+      "grad_norm": 0.2504301965236664,
+      "learning_rate": 9.9228915070518e-06,
+      "loss": 0.8382,
+      "step": 220
+    },
+    {
+      "epoch": 0.08855520261815382,
+      "grad_norm": 0.2495463490486145,
+      "learning_rate": 9.911599764737632e-06,
+      "loss": 0.8393,
+      "step": 230
+    },
+    {
+      "epoch": 0.09240542881894312,
+      "grad_norm": 0.24468041956424713,
+      "learning_rate": 9.899544079783606e-06,
+      "loss": 0.8341,
+      "step": 240
+    },
+    {
+      "epoch": 0.0962556550197324,
+      "grad_norm": 0.29563626646995544,
+      "learning_rate": 9.886726327312338e-06,
+      "loss": 0.8504,
+      "step": 250
+    },
+    {
+      "epoch": 0.1001058812205217,
+      "grad_norm": 0.24205781519412994,
+      "learning_rate": 9.87314850097724e-06,
+      "loss": 0.8475,
+      "step": 260
+    },
+    {
+      "epoch": 0.103956107421311,
+      "grad_norm": 0.24268938601016998,
+      "learning_rate": 9.858812712652452e-06,
+      "loss": 0.835,
+      "step": 270
+    },
+    {
+      "epoch": 0.10780633362210029,
+      "grad_norm": 0.23990951478481293,
+      "learning_rate": 9.843721192104341e-06,
+      "loss": 0.8348,
+      "step": 280
+    },
+    {
+      "epoch": 0.11165655982288959,
+      "grad_norm": 0.25300315022468567,
+      "learning_rate": 9.827876286644698e-06,
+      "loss": 0.852,
+      "step": 290
+    },
+    {
+      "epoch": 0.11550678602367889,
+      "grad_norm": 0.22314424812793732,
+      "learning_rate": 9.811280460765643e-06,
+      "loss": 0.8133,
+      "step": 300
+    },
+    {
+      "epoch": 0.11935701222446819,
+      "grad_norm": 0.2356373518705368,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.8265,
+      "step": 310
+    },
+    {
+      "epoch": 0.12320723842525748,
+      "grad_norm": 0.25164154171943665,
+      "learning_rate": 9.775846489301279e-06,
+      "loss": 0.8261,
+      "step": 320
+    },
+    {
+      "epoch": 0.12705746462604678,
+      "grad_norm": 0.26003730297088623,
+      "learning_rate": 9.757013855061157e-06,
+      "loss": 0.8265,
+      "step": 330
+    },
+    {
+      "epoch": 0.13090769082683607,
+      "grad_norm": 0.23910772800445557,
+      "learning_rate": 9.737441322234758e-06,
+      "loss": 0.8373,
+      "step": 340
+    },
+    {
+      "epoch": 0.13475791702762538,
+      "grad_norm": 0.2616579830646515,
+      "learning_rate": 9.717131935103605e-06,
+      "loss": 0.8374,
+      "step": 350
+    },
+    {
+      "epoch": 0.13860814322841467,
+      "grad_norm": 0.24132589995861053,
+      "learning_rate": 9.696088852558398e-06,
+      "loss": 0.8239,
+      "step": 360
+    },
+    {
+      "epoch": 0.14245836942920398,
+      "grad_norm": 0.25307944416999817,
+      "learning_rate": 9.674315347607686e-06,
+      "loss": 0.8503,
+      "step": 370
+    },
+    {
+      "epoch": 0.14630859562999327,
+      "grad_norm": 0.23904776573181152,
+      "learning_rate": 9.651814806868796e-06,
+      "loss": 0.8101,
+      "step": 380
+    },
+    {
+      "epoch": 0.15015882183078255,
+      "grad_norm": 0.23685136437416077,
+      "learning_rate": 9.62859073004107e-06,
+      "loss": 0.817,
+      "step": 390
+    },
+    {
+      "epoch": 0.15400904803157187,
+      "grad_norm": 0.23827877640724182,
+      "learning_rate": 9.604646729361547e-06,
+      "loss": 0.82,
+      "step": 400
+    },
+    {
+      "epoch": 0.15785927423236115,
+      "grad_norm": 0.23255571722984314,
+      "learning_rate": 9.579986529043099e-06,
+      "loss": 0.8097,
+      "step": 410
+    },
+    {
+      "epoch": 0.16170950043315044,
+      "grad_norm": 0.2344449758529663,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.826,
+      "step": 420
+    },
+    {
+      "epoch": 0.16555972663393975,
+      "grad_norm": 0.2431863248348236,
+      "learning_rate": 9.528532982727281e-06,
+      "loss": 0.8155,
+      "step": 430
+    },
+    {
+      "epoch": 0.16940995283472904,
+      "grad_norm": 0.24152682721614838,
+      "learning_rate": 9.501747639735018e-06,
+      "loss": 0.813,
+      "step": 440
+    },
+    {
+      "epoch": 0.17326017903551832,
+      "grad_norm": 0.2492619901895523,
+      "learning_rate": 9.474262101869272e-06,
+      "loss": 0.8034,
+      "step": 450
+    },
+    {
+      "epoch": 0.17711040523630764,
+      "grad_norm": 0.23506779968738556,
+      "learning_rate": 9.446080644188141e-06,
+      "loss": 0.8085,
+      "step": 460
+    },
+    {
+      "epoch": 0.18096063143709692,
+      "grad_norm": 0.24165698885917664,
+      "learning_rate": 9.417207649992017e-06,
+      "loss": 0.8167,
+      "step": 470
+    },
+    {
+      "epoch": 0.18481085763788624,
+      "grad_norm": 0.2382257878780365,
+      "learning_rate": 9.387647610141816e-06,
+      "loss": 0.8152,
+      "step": 480
+    },
+    {
+      "epoch": 0.18866108383867553,
+      "grad_norm": 0.22803795337677002,
+      "learning_rate": 9.35740512236047e-06,
+      "loss": 0.802,
+      "step": 490
+    },
+    {
+      "epoch": 0.1925113100394648,
+      "grad_norm": 0.22222229838371277,
+      "learning_rate": 9.326484890517809e-06,
+      "loss": 0.8073,
+      "step": 500
+    },
+    {
+      "epoch": 0.19636153624025413,
+      "grad_norm": 0.2322579324245453,
+      "learning_rate": 9.294891723898926e-06,
+      "loss": 0.8021,
+      "step": 510
+    },
+    {
+      "epoch": 0.2002117624410434,
+      "grad_norm": 0.24413122236728668,
+      "learning_rate": 9.262630536456153e-06,
+      "loss": 0.8319,
+      "step": 520
+    },
+    {
+      "epoch": 0.2040619886418327,
+      "grad_norm": 0.2532041072845459,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.8035,
+      "step": 530
+    },
+    {
+      "epoch": 0.207912214842622,
+      "grad_norm": 0.25332382321357727,
+      "learning_rate": 9.196124273642422e-06,
+      "loss": 0.8173,
+      "step": 540
+    },
+    {
+      "epoch": 0.2117624410434113,
+      "grad_norm": 0.22490622103214264,
+      "learning_rate": 9.161889542552842e-06,
+      "loss": 0.819,
+      "step": 550
+    },
+    {
+      "epoch": 0.21561266724420058,
+      "grad_norm": 0.2185283750295639,
+      "learning_rate": 9.127007477593193e-06,
+      "loss": 0.8236,
+      "step": 560
+    },
+    {
+      "epoch": 0.2194628934449899,
+      "grad_norm": 0.2579803466796875,
+      "learning_rate": 9.091483504265974e-06,
+      "loss": 0.8319,
+      "step": 570
+    },
+    {
+      "epoch": 0.22331311964577918,
+      "grad_norm": 0.2508712708950043,
+      "learning_rate": 9.055323147915116e-06,
+      "loss": 0.812,
+      "step": 580
+    },
+    {
+      "epoch": 0.2271633458465685,
+      "grad_norm": 0.24880196154117584,
+      "learning_rate": 9.018532032866585e-06,
+      "loss": 0.8221,
+      "step": 590
+    },
+    {
+      "epoch": 0.23101357204735778,
+      "grad_norm": 0.23069332540035248,
+      "learning_rate": 8.981115881553579e-06,
+      "loss": 0.8128,
+      "step": 600
+    },
+    {
+      "epoch": 0.23486379824814707,
+      "grad_norm": 0.25827839970588684,
+      "learning_rate": 8.943080513626476e-06,
+      "loss": 0.8323,
+      "step": 610
+    },
+    {
+      "epoch": 0.23871402444893638,
+      "grad_norm": 0.23462145030498505,
+      "learning_rate": 8.904431845047641e-06,
+      "loss": 0.8183,
+      "step": 620
+    },
+    {
+      "epoch": 0.24256425064972567,
+      "grad_norm": 0.2504104971885681,
+      "learning_rate": 8.86517588717128e-06,
+      "loss": 0.834,
+      "step": 630
+    },
+    {
+      "epoch": 0.24641447685051496,
+      "grad_norm": 0.23117466270923615,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.8231,
+      "step": 640
+    },
+    {
+      "epoch": 0.25026470305130427,
+      "grad_norm": 0.2247200757265091,
+      "learning_rate": 8.784866620277314e-06,
+      "loss": 0.7805,
+      "step": 650
+    },
+    {
+      "epoch": 0.25411492925209356,
+      "grad_norm": 0.2575814723968506,
+      "learning_rate": 8.743825802439018e-06,
+      "loss": 0.8164,
+      "step": 660
+    },
+    {
+      "epoch": 0.25796515545288284,
+      "grad_norm": 0.25640708208084106,
+      "learning_rate": 8.702202675718956e-06,
+      "loss": 0.8221,
+      "step": 670
+    },
+    {
+      "epoch": 0.26181538165367213,
+      "grad_norm": 0.23116858303546906,
+      "learning_rate": 8.660003714113959e-06,
+      "loss": 0.7917,
+      "step": 680
+    },
+    {
+      "epoch": 0.26566560785446147,
+      "grad_norm": 0.24693791568279266,
+      "learning_rate": 8.617235481185323e-06,
+      "loss": 0.8167,
+      "step": 690
+    },
+    {
+      "epoch": 0.26951583405525076,
+      "grad_norm": 0.24754402041435242,
+      "learning_rate": 8.573904629037931e-06,
+      "loss": 0.822,
+      "step": 700
+    },
+    {
+      "epoch": 0.27336606025604004,
+      "grad_norm": 0.23752228915691376,
+      "learning_rate": 8.530017897285585e-06,
+      "loss": 0.8016,
+      "step": 710
+    },
+    {
+      "epoch": 0.27721628645682933,
+      "grad_norm": 0.21996991336345673,
+      "learning_rate": 8.485582112002746e-06,
+      "loss": 0.8056,
+      "step": 720
+    },
+    {
+      "epoch": 0.2810665126576186,
+      "grad_norm": 0.2329818308353424,
+      "learning_rate": 8.440604184662811e-06,
+      "loss": 0.8115,
+      "step": 730
+    },
+    {
+      "epoch": 0.28491673885840796,
+      "grad_norm": 0.23024459183216095,
+      "learning_rate": 8.395091111063118e-06,
+      "loss": 0.8001,
+      "step": 740
+    },
+    {
+      "epoch": 0.28876696505919724,
+      "grad_norm": 0.2226322442293167,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.7999,
+      "step": 750
+    },
+    {
+      "epoch": 0.29261719125998653,
+      "grad_norm": 0.2354160100221634,
+      "learning_rate": 8.302487923351842e-06,
+      "loss": 0.7922,
+      "step": 760
+    },
+    {
+      "epoch": 0.2964674174607758,
+      "grad_norm": 0.23037636280059814,
+      "learning_rate": 8.255412212597021e-06,
+      "loss": 0.8097,
+      "step": 770
+    },
+    {
+      "epoch": 0.3003176436615651,
+      "grad_norm": 0.2321627140045166,
+      "learning_rate": 8.207830160055688e-06,
+      "loss": 0.8065,
+      "step": 780
+    },
+    {
+      "epoch": 0.3041678698623544,
+      "grad_norm": 0.2403370589017868,
+      "learning_rate": 8.159749166566786e-06,
+      "loss": 0.7905,
+      "step": 790
+    },
+    {
+      "epoch": 0.30801809606314373,
+      "grad_norm": 0.22592222690582275,
+      "learning_rate": 8.11117671057377e-06,
+      "loss": 0.8183,
+      "step": 800
+    },
+    {
+      "epoch": 0.311868322263933,
+      "grad_norm": 0.23806729912757874,
+      "learning_rate": 8.062120346961404e-06,
+      "loss": 0.8109,
+      "step": 810
+    },
+    {
+      "epoch": 0.3157185484647223,
+      "grad_norm": 0.22496743500232697,
+      "learning_rate": 8.012587705880705e-06,
+      "loss": 0.8025,
+      "step": 820
+    },
+    {
+      "epoch": 0.3195687746655116,
+      "grad_norm": 0.23336046934127808,
+      "learning_rate": 7.962586491562146e-06,
+      "loss": 0.7905,
+      "step": 830
+    },
+    {
+      "epoch": 0.3234190008663009,
+      "grad_norm": 0.24823102355003357,
+      "learning_rate": 7.912124481117357e-06,
+      "loss": 0.7957,
+      "step": 840
+    },
+    {
+      "epoch": 0.3272692270670902,
+      "grad_norm": 0.2368079572916031,
+      "learning_rate": 7.861209523329494e-06,
+      "loss": 0.8058,
+      "step": 850
+    },
+    {
+      "epoch": 0.3311194532678795,
+      "grad_norm": 0.23666667938232422,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.7889,
+      "step": 860
+    },
+    {
+      "epoch": 0.3349696794686688,
+      "grad_norm": 0.23386090993881226,
+      "learning_rate": 7.758052511879035e-06,
+      "loss": 0.8006,
+      "step": 870
+    },
+    {
+      "epoch": 0.3388199056694581,
+      "grad_norm": 0.2251926064491272,
+      "learning_rate": 7.705826503098642e-06,
+      "loss": 0.7947,
+      "step": 880
+    },
+    {
+      "epoch": 0.34267013187024736,
+      "grad_norm": 0.230204775929451,
+      "learning_rate": 7.653179634243979e-06,
+      "loss": 0.7955,
+      "step": 890
+    },
+    {
+      "epoch": 0.34652035807103665,
+      "grad_norm": 0.22313480079174042,
+      "learning_rate": 7.600120093927696e-06,
+      "loss": 0.8001,
+      "step": 900
+    },
+    {
+      "epoch": 0.350370584271826,
+      "grad_norm": 0.23587562143802643,
+      "learning_rate": 7.546656134948729e-06,
+      "loss": 0.7902,
+      "step": 910
+    },
+    {
+      "epoch": 0.3542208104726153,
+      "grad_norm": 0.24195075035095215,
+      "learning_rate": 7.492796073008661e-06,
+      "loss": 0.7888,
+      "step": 920
+    },
+    {
+      "epoch": 0.35807103667340456,
+      "grad_norm": 0.24929076433181763,
+      "learning_rate": 7.438548285418326e-06,
+      "loss": 0.7746,
+      "step": 930
+    },
+    {
+      "epoch": 0.36192126287419385,
+      "grad_norm": 0.2498975545167923,
+      "learning_rate": 7.3839212097948e-06,
+      "loss": 0.8087,
+      "step": 940
+    },
+    {
+      "epoch": 0.36577148907498314,
+      "grad_norm": 0.2354947328567505,
+      "learning_rate": 7.328923342749039e-06,
+      "loss": 0.8237,
+      "step": 950
+    },
+    {
+      "epoch": 0.3696217152757725,
+      "grad_norm": 0.2525039315223694,
+      "learning_rate": 7.273563238564319e-06,
+      "loss": 0.8071,
+      "step": 960
+    },
+    {
+      "epoch": 0.37347194147656176,
+      "grad_norm": 0.2385227233171463,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.7881,
+      "step": 970
+    },
+    {
+      "epoch": 0.37732216767735105,
+      "grad_norm": 0.2302110195159912,
+      "learning_rate": 7.161790816280858e-06,
+      "loss": 0.7933,
+      "step": 980
+    },
+    {
+      "epoch": 0.38117239387814034,
+      "grad_norm": 0.24021956324577332,
+      "learning_rate": 7.1053958830920125e-06,
+      "loss": 0.7799,
+      "step": 990
+    },
+    {
+      "epoch": 0.3850226200789296,
+      "grad_norm": 0.23409691452980042,
+      "learning_rate": 7.0486734798799715e-06,
+      "loss": 0.7963,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3888728462797189,
+      "grad_norm": 0.2414340227842331,
+      "learning_rate": 6.991632429159708e-06,
+      "loss": 0.8189,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39272307248050825,
+      "grad_norm": 0.242491215467453,
+      "learning_rate": 6.934281603008132e-06,
+      "loss": 0.8098,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39657329868129754,
+      "grad_norm": 0.23040764033794403,
+      "learning_rate": 6.876629921684143e-06,
+      "loss": 0.8067,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4004235248820868,
+      "grad_norm": 0.23939909040927887,
+      "learning_rate": 6.818686352241199e-06,
+      "loss": 0.798,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4042737510828761,
+      "grad_norm": 0.2218620777130127,
+      "learning_rate": 6.760459907132573e-06,
+      "loss": 0.7946,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4081239772836654,
+      "grad_norm": 0.21917887032032013,
+      "learning_rate": 6.701959642809589e-06,
+      "loss": 0.8006,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41197420348445474,
+      "grad_norm": 0.23311498761177063,
+      "learning_rate": 6.643194658312987e-06,
+      "loss": 0.7937,
+      "step": 1070
+    },
+    {
+      "epoch": 0.415824429685244,
+      "grad_norm": 0.23265351355075836,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.8161,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4196746558860333,
+      "grad_norm": 0.26503854990005493,
+      "learning_rate": 6.5249071294110735e-06,
+      "loss": 0.7963,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4235248820868226,
+      "grad_norm": 0.24315175414085388,
+      "learning_rate": 6.465402983265274e-06,
+      "loss": 0.8108,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4273751082876119,
+      "grad_norm": 0.24192595481872559,
+      "learning_rate": 6.405670910603243e-06,
+      "loss": 0.8066,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43122533448840117,
+      "grad_norm": 0.253692626953125,
+      "learning_rate": 6.345720202059291e-06,
+      "loss": 0.8029,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4350755606891905,
+      "grad_norm": 0.2414553314447403,
+      "learning_rate": 6.285560182274008e-06,
+      "loss": 0.7815,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4389257868899798,
+      "grad_norm": 0.23615378141403198,
+      "learning_rate": 6.225200208443936e-06,
+      "loss": 0.8037,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4427760130907691,
+      "grad_norm": 0.2203551083803177,
+      "learning_rate": 6.16464966886616e-06,
+      "loss": 0.7936,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44662623929155837,
+      "grad_norm": 0.26320552825927734,
+      "learning_rate": 6.103917981478065e-06,
+      "loss": 0.8167,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45047646549234766,
+      "grad_norm": 0.22850251197814941,
+      "learning_rate": 6.043014592392489e-06,
+      "loss": 0.828,
+      "step": 1170
+    },
+    {
+      "epoch": 0.454326691693137,
+      "grad_norm": 0.227044016122818,
+      "learning_rate": 5.981948974428485e-06,
+      "loss": 0.8025,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4581769178939263,
+      "grad_norm": 0.2402268350124359,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.782,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46202714409471557,
+      "grad_norm": 0.24159236252307892,
+      "learning_rate": 5.8593690678282355e-06,
+      "loss": 0.7922,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46587737029550486,
+      "grad_norm": 0.24446304142475128,
+      "learning_rate": 5.797873845081301e-06,
+      "loss": 0.7969,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46972759649629414,
+      "grad_norm": 0.22982114553451538,
+      "learning_rate": 5.736254522269075e-06,
+      "loss": 0.813,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47357782269708343,
+      "grad_norm": 0.23523172736167908,
+      "learning_rate": 5.6745206835658386e-06,
+      "loss": 0.8146,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47742804889787277,
+      "grad_norm": 0.22839456796646118,
+      "learning_rate": 5.612681930957495e-06,
+      "loss": 0.7859,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48127827509866206,
+      "grad_norm": 0.22501277923583984,
+      "learning_rate": 5.550747882748095e-06,
+      "loss": 0.7848,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48512850129945134,
+      "grad_norm": 0.2500593960285187,
+      "learning_rate": 5.4887281720638195e-06,
+      "loss": 0.79,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48897872750024063,
+      "grad_norm": 0.2635622024536133,
+      "learning_rate": 5.42663244535466e-06,
+      "loss": 0.7762,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4928289537010299,
+      "grad_norm": 0.2632523477077484,
+      "learning_rate": 5.3644703608940205e-06,
+      "loss": 0.7844,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49667917990181926,
+      "grad_norm": 0.23444679379463196,
+      "learning_rate": 5.302251587276491e-06,
+      "loss": 0.7947,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5005294061026085,
+      "grad_norm": 0.2246859222650528,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.7942,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5043796323033978,
+      "grad_norm": 0.23471280932426453,
+      "learning_rate": 5.177682689530615e-06,
+      "loss": 0.8098,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5082298585041871,
+      "grad_norm": 0.2407495528459549,
+      "learning_rate": 5.11535194065619e-06,
+      "loss": 0.7768,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5120800847049765,
+      "grad_norm": 0.26251131296157837,
+      "learning_rate": 5.053003250119113e-06,
+      "loss": 0.7887,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5159303109057657,
+      "grad_norm": 0.23856370151042938,
+      "learning_rate": 4.9906463155383925e-06,
+      "loss": 0.7948,
+      "step": 1340
+    },
+    {
+      "epoch": 0.519780537106555,
+      "grad_norm": 0.2380622774362564,
+      "learning_rate": 4.9282908358153046e-06,
+      "loss": 0.7865,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5236307633073443,
+      "grad_norm": 0.2565951645374298,
+      "learning_rate": 4.8659465096248345e-06,
+      "loss": 0.8052,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5274809895081336,
+      "grad_norm": 0.24584832787513733,
+      "learning_rate": 4.803623033907169e-06,
+      "loss": 0.8042,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5313312157089229,
+      "grad_norm": 0.22715476155281067,
+      "learning_rate": 4.741330102359441e-06,
+      "loss": 0.8094,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5351814419097122,
+      "grad_norm": 0.24714422225952148,
+      "learning_rate": 4.679077403927996e-06,
+      "loss": 0.7888,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5390316681105015,
+      "grad_norm": 0.23103509843349457,
+      "learning_rate": 4.616874621301373e-06,
+      "loss": 0.7884,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5428818943112907,
+      "grad_norm": 0.24038435518741608,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.8031,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5467321205120801,
+      "grad_norm": 0.24905183911323547,
+      "learning_rate": 4.492657493892822e-06,
+      "loss": 0.7888,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5505823467128694,
+      "grad_norm": 0.28134477138519287,
+      "learning_rate": 4.430662469650999e-06,
+      "loss": 0.8039,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5544325729136587,
+      "grad_norm": 0.2530960142612457,
+      "learning_rate": 4.368755999289118e-06,
+      "loss": 0.8053,
+      "step": 1440
+    },
+    {
+      "epoch": 0.558282799114448,
+      "grad_norm": 0.26474103331565857,
+      "learning_rate": 4.306947711643949e-06,
+      "loss": 0.7913,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5621330253152372,
+      "grad_norm": 0.2411193549633026,
+      "learning_rate": 4.245247220281068e-06,
+      "loss": 0.7937,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5659832515160266,
+      "grad_norm": 0.23446902632713318,
+      "learning_rate": 4.183664121999587e-06,
+      "loss": 0.791,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5698334777168159,
+      "grad_norm": 0.2463819980621338,
+      "learning_rate": 4.122207995339474e-06,
+      "loss": 0.7879,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5736837039176051,
+      "grad_norm": 0.2318504899740219,
+      "learning_rate": 4.060888399091734e-06,
+      "loss": 0.7929,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5775339301183945,
+      "grad_norm": 0.23570016026496887,
+      "learning_rate": 3.999714870811639e-06,
+      "loss": 0.8007,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5813841563191837,
+      "grad_norm": 0.23344284296035767,
+      "learning_rate": 3.938696925335275e-06,
+      "loss": 0.7947,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5852343825199731,
+      "grad_norm": 0.23675939440727234,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.802,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5890846087207623,
+      "grad_norm": 0.24015510082244873,
+      "learning_rate": 3.817165719666382e-06,
+      "loss": 0.8049,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5929348349215516,
+      "grad_norm": 0.2332915961742401,
+      "learning_rate": 3.756671362249841e-06,
+      "loss": 0.8074,
+      "step": 1540
+    },
+    {
+      "epoch": 0.596785061122341,
+      "grad_norm": 0.2314557284116745,
+      "learning_rate": 3.6963703902488907e-06,
+      "loss": 0.7944,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6006352873231302,
+      "grad_norm": 0.22771933674812317,
+      "learning_rate": 3.6362721827835655e-06,
+      "loss": 0.8131,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6044855135239195,
+      "grad_norm": 0.24463161826133728,
+      "learning_rate": 3.576386087436212e-06,
+      "loss": 0.8062,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6083357397247088,
+      "grad_norm": 0.2524418830871582,
+      "learning_rate": 3.5167214187975913e-06,
+      "loss": 0.7968,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6121859659254981,
+      "grad_norm": 0.20595966279506683,
+      "learning_rate": 3.457287457018086e-06,
+      "loss": 0.7986,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6160361921262875,
+      "grad_norm": 0.231485977768898,
+      "learning_rate": 3.3980934463643003e-06,
+      "loss": 0.7856,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6198864183270767,
+      "grad_norm": 0.21625636518001556,
+      "learning_rate": 3.3391485937812053e-06,
+      "loss": 0.769,
+      "step": 1610
+    },
+    {
+      "epoch": 0.623736644527866,
+      "grad_norm": 0.24160811305046082,
+      "learning_rate": 3.280462067460115e-06,
+      "loss": 0.7681,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6275868707286553,
+      "grad_norm": 0.23286288976669312,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.7982,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6314370969294446,
+      "grad_norm": 0.21974439918994904,
+      "learning_rate": 3.163900464051089e-06,
+      "loss": 0.7895,
+      "step": 1640
+    },
+    {
+      "epoch": 0.635287323130234,
+      "grad_norm": 0.24553722143173218,
+      "learning_rate": 3.1060435167748826e-06,
+      "loss": 0.7973,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6391375493310232,
+      "grad_norm": 0.24239341914653778,
+      "learning_rate": 3.048481152564253e-06,
+      "loss": 0.7788,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6429877755318125,
+      "grad_norm": 0.2442914843559265,
+      "learning_rate": 2.9912223245804017e-06,
+      "loss": 0.7999,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6468380017326018,
+      "grad_norm": 0.24031966924667358,
+      "learning_rate": 2.934275938772981e-06,
+      "loss": 0.7969,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6506882279333911,
+      "grad_norm": 0.25379085540771484,
+      "learning_rate": 2.8776508524948676e-06,
+      "loss": 0.7869,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6545384541341804,
+      "grad_norm": 0.21629001200199127,
+      "learning_rate": 2.821355873124503e-06,
+      "loss": 0.7851,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6583886803349697,
+      "grad_norm": 0.24450668692588806,
+      "learning_rate": 2.7653997566960144e-06,
+      "loss": 0.7989,
+      "step": 1710
+    },
+    {
+      "epoch": 0.662238906535759,
+      "grad_norm": 0.25176525115966797,
+      "learning_rate": 2.709791206537309e-06,
+      "loss": 0.789,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6660891327365482,
+      "grad_norm": 0.23535434901714325,
+      "learning_rate": 2.6545388719163747e-06,
+      "loss": 0.7954,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6699393589373376,
+      "grad_norm": 0.2267545461654663,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.7927,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6737895851381268,
+      "grad_norm": 0.24057923257350922,
+      "learning_rate": 2.545137167997007e-06,
+      "loss": 0.7892,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6776398113389162,
+      "grad_norm": 0.22840392589569092,
+      "learning_rate": 2.4910048148705972e-06,
+      "loss": 0.7776,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6814900375397055,
+      "grad_norm": 0.24809467792510986,
+      "learning_rate": 2.4372627069793327e-06,
+      "loss": 0.8213,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6853402637404947,
+      "grad_norm": 0.2365812063217163,
+      "learning_rate": 2.3839192032876486e-06,
+      "loss": 0.8067,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6891904899412841,
+      "grad_norm": 0.23032653331756592,
+      "learning_rate": 2.3309826007617047e-06,
+      "loss": 0.8176,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6930407161420733,
+      "grad_norm": 0.2400519996881485,
+      "learning_rate": 2.278461133078881e-06,
+      "loss": 0.7821,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6968909423428626,
+      "grad_norm": 0.2368413358926773,
+      "learning_rate": 2.2263629693471258e-06,
+      "loss": 0.7717,
+      "step": 1810
+    },
+    {
+      "epoch": 0.700741168543652,
+      "grad_norm": 0.22876349091529846,
+      "learning_rate": 2.1746962128343414e-06,
+      "loss": 0.7972,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7045913947444412,
+      "grad_norm": 0.22780956327915192,
+      "learning_rate": 2.1234688997080183e-06,
+      "loss": 0.7852,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7084416209452306,
+      "grad_norm": 0.25013190507888794,
+      "learning_rate": 2.0726889977853053e-06,
+      "loss": 0.7901,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7122918471460198,
+      "grad_norm": 0.24060063064098358,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.8037,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7161420733468091,
+      "grad_norm": 0.23506894707679749,
+      "learning_rate": 1.9725029496425814e-06,
+      "loss": 0.7939,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7199922995475985,
+      "grad_norm": 0.24514800310134888,
+      "learning_rate": 1.923112386205727e-06,
+      "loss": 0.7925,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7238425257483877,
+      "grad_norm": 0.21959026157855988,
+      "learning_rate": 1.874200397115079e-06,
+      "loss": 0.7979,
+      "step": 1880
+    },
+    {
+      "epoch": 0.727692751949177,
+      "grad_norm": 0.2392028570175171,
+      "learning_rate": 1.825774590065857e-06,
+      "loss": 0.7834,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7315429781499663,
+      "grad_norm": 0.24447238445281982,
+      "learning_rate": 1.7778424971332813e-06,
+      "loss": 0.7893,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7353932043507556,
+      "grad_norm": 0.22846662998199463,
+      "learning_rate": 1.7304115736010407e-06,
+      "loss": 0.7891,
+      "step": 1910
+    },
+    {
+      "epoch": 0.739243430551545,
+      "grad_norm": 0.23552203178405762,
+      "learning_rate": 1.6834891968017108e-06,
+      "loss": 0.7804,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7430936567523342,
+      "grad_norm": 0.24262794852256775,
+      "learning_rate": 1.6370826649692884e-06,
+      "loss": 0.8004,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7469438829531235,
+      "grad_norm": 0.2340243011713028,
+      "learning_rate": 1.5911991961040463e-06,
+      "loss": 0.8087,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7507941091539128,
+      "grad_norm": 0.24299858510494232,
+      "learning_rate": 1.545845926849851e-06,
+      "loss": 0.7953,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7546443353547021,
+      "grad_norm": 0.25273993611335754,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.8127,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7584945615554913,
+      "grad_norm": 0.2365594059228897,
+      "learning_rate": 1.4567581203207238e-06,
+      "loss": 0.8005,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7623447877562807,
+      "grad_norm": 0.24935095012187958,
+      "learning_rate": 1.413037439625593e-06,
+      "loss": 0.8006,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76619501395707,
+      "grad_norm": 0.2501228153705597,
+      "learning_rate": 1.3698746695458847e-06,
+      "loss": 0.7937,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7700452401578592,
+      "grad_norm": 0.23038862645626068,
+      "learning_rate": 1.3272765235521745e-06,
+      "loss": 0.7853,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7738954663586486,
+      "grad_norm": 0.22457218170166016,
+      "learning_rate": 1.2852496272942877e-06,
+      "loss": 0.7931,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7777456925594378,
+      "grad_norm": 0.21993689239025116,
+      "learning_rate": 1.2438005175707396e-06,
+      "loss": 0.7813,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7815959187602272,
+      "grad_norm": 0.23017768561840057,
+      "learning_rate": 1.2029356413120297e-06,
+      "loss": 0.7672,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7854461449610165,
+      "grad_norm": 0.23289257287979126,
+      "learning_rate": 1.1626613545778786e-06,
+      "loss": 0.7842,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7892963711618057,
+      "grad_norm": 0.22083888947963715,
+      "learning_rate": 1.1229839215686294e-06,
+      "loss": 0.7879,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7931465973625951,
+      "grad_norm": 0.2122008353471756,
+      "learning_rate": 1.0839095136509188e-06,
+      "loss": 0.7745,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7969968235633843,
+      "grad_norm": 0.2593449652194977,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.796,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8008470497641736,
+      "grad_norm": 0.22876858711242676,
+      "learning_rate": 1.007593988643402e-06,
+      "loss": 0.7755,
+      "step": 2080
+    },
+    {
+      "epoch": 0.804697275964963,
+      "grad_norm": 0.23252536356449127,
+      "learning_rate": 9.703647415524613e-07,
+      "loss": 0.7981,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8085475021657522,
+      "grad_norm": 0.22701573371887207,
+      "learning_rate": 9.337622577045485e-07,
+      "loss": 0.7866,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8123977283665416,
+      "grad_norm": 0.22942154109477997,
+      "learning_rate": 8.977922301934588e-07,
+      "loss": 0.7949,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8162479545673308,
+      "grad_norm": 0.24092769622802734,
+      "learning_rate": 8.624602537417009e-07,
+      "loss": 0.7982,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8200981807681201,
+      "grad_norm": 0.2467198222875595,
+      "learning_rate": 8.277718238303123e-07,
+      "loss": 0.786,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8239484069689095,
+      "grad_norm": 0.2407129555940628,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.7877,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8277986331696987,
+      "grad_norm": 0.24136090278625488,
+      "learning_rate": 7.603470842324223e-07,
+      "loss": 0.7813,
+      "step": 2150
+    },
+    {
+      "epoch": 0.831648859370488,
+      "grad_norm": 0.22663679718971252,
+      "learning_rate": 7.276212616857575e-07,
+      "loss": 0.798,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8354990855712773,
+      "grad_norm": 0.22803762555122375,
+      "learning_rate": 6.955599583279837e-07,
+      "loss": 0.7993,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8393493117720666,
+      "grad_norm": 0.23298944532871246,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.7893,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8431995379728559,
+      "grad_norm": 0.22421827912330627,
+      "learning_rate": 6.334507521075711e-07,
+      "loss": 0.7975,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8470497641736452,
+      "grad_norm": 0.25587308406829834,
+      "learning_rate": 6.034125096149179e-07,
+      "loss": 0.79,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8508999903744345,
+      "grad_norm": 0.2331983596086502,
+      "learning_rate": 5.740581055485722e-07,
+      "loss": 0.7628,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8547502165752238,
+      "grad_norm": 0.22114115953445435,
+      "learning_rate": 5.453921056471961e-07,
+      "loss": 0.7968,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8586004427760131,
+      "grad_norm": 0.24111635982990265,
+      "learning_rate": 5.174189685761288e-07,
+      "loss": 0.7899,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8624506689768023,
+      "grad_norm": 0.23629552125930786,
+      "learning_rate": 4.901430452338979e-07,
+      "loss": 0.8021,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8663008951775917,
+      "grad_norm": 0.2360439896583557,
+      "learning_rate": 4.6356857807548286e-07,
+      "loss": 0.7858,
+      "step": 2250
+    },
+    {
+      "epoch": 0.870151121378381,
+      "grad_norm": 0.23215076327323914,
+      "learning_rate": 4.376997004524541e-07,
+      "loss": 0.7748,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8740013475791703,
+      "grad_norm": 0.243072509765625,
+      "learning_rate": 4.12540435970073e-07,
+      "loss": 0.7862,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8778515737799596,
+      "grad_norm": 0.2262149453163147,
+      "learning_rate": 3.880946978614703e-07,
+      "loss": 0.7942,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8817017999807488,
+      "grad_norm": 0.23143689334392548,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.7992,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8855520261815382,
+      "grad_norm": 0.22167249023914337,
+      "learning_rate": 3.41358898202781e-07,
+      "loss": 0.7756,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8894022523823275,
+      "grad_norm": 0.22288759052753448,
+      "learning_rate": 3.190761058667741e-07,
+      "loss": 0.7875,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8932524785831167,
+      "grad_norm": 0.2282417267560959,
+      "learning_rate": 2.9752137720206463e-07,
+      "loss": 0.7954,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8971027047839061,
+      "grad_norm": 0.23069989681243896,
+      "learning_rate": 2.7669806479785067e-07,
+      "loss": 0.7988,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9009529309846953,
+      "grad_norm": 0.24257265031337738,
+      "learning_rate": 2.566094074799763e-07,
+      "loss": 0.7936,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9048031571854847,
+      "grad_norm": 0.22323960065841675,
+      "learning_rate": 2.3725852980716557e-07,
+      "loss": 0.7737,
+      "step": 2350
+    },
+    {
+      "epoch": 0.908653383386274,
+      "grad_norm": 0.22651095688343048,
+      "learning_rate": 2.1864844158503983e-07,
+      "loss": 0.7882,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9125036095870632,
+      "grad_norm": 0.252742201089859,
+      "learning_rate": 2.007820373979702e-07,
+      "loss": 0.7796,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9163538357878526,
+      "grad_norm": 0.23900404572486877,
+      "learning_rate": 1.8366209615886532e-07,
+      "loss": 0.7876,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9202040619886418,
+      "grad_norm": 0.23098358511924744,
+      "learning_rate": 1.6729128067693613e-07,
+      "loss": 0.7984,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9240542881894311,
+      "grad_norm": 0.25214311480522156,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.7992,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9279045143902205,
+      "grad_norm": 0.22786889970302582,
+      "learning_rate": 1.3680709523610058e-07,
+      "loss": 0.7825,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9317547405910097,
+      "grad_norm": 0.22502072155475616,
+      "learning_rate": 1.2269846674030418e-07,
+      "loss": 0.8036,
+      "step": 2420
+    },
+    {
+      "epoch": 0.935604966791799,
+      "grad_norm": 0.2321796864271164,
+      "learning_rate": 1.0934844619042595e-07,
+      "loss": 0.8008,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9394551929925883,
+      "grad_norm": 0.2391970157623291,
+      "learning_rate": 9.675911002803961e-08,
+      "loss": 0.79,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9433054191933776,
+      "grad_norm": 0.2453995794057846,
+      "learning_rate": 8.493241637904381e-08,
+      "loss": 0.7865,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9471556453941669,
+      "grad_norm": 0.23847082257270813,
+      "learning_rate": 7.387020474909689e-08,
+      "loss": 0.8071,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9510058715949562,
+      "grad_norm": 0.24866542220115662,
+      "learning_rate": 6.357419573750856e-08,
+      "loss": 0.8014,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9548560977957455,
+      "grad_norm": 0.23370376229286194,
+      "learning_rate": 5.4045990769613365e-08,
+      "loss": 0.7844,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9587063239965348,
+      "grad_norm": 0.2773272693157196,
+      "learning_rate": 4.5287071847694434e-08,
+      "loss": 0.79,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9625565501973241,
+      "grad_norm": 0.2278584986925125,
+      "learning_rate": 3.7298801320467817e-08,
+      "loss": 0.7834,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9664067763981133,
+      "grad_norm": 0.23154611885547638,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.777,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9702570025989027,
+      "grad_norm": 0.23624077439308167,
+      "learning_rate": 2.3639055324410355e-08,
+      "loss": 0.7992,
+      "step": 2520
+    },
+    {
+      "epoch": 0.974107228799692,
+      "grad_norm": 0.22888708114624023,
+      "learning_rate": 1.7969704471369477e-08,
+      "loss": 0.7827,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9779574550004813,
+      "grad_norm": 0.22431617975234985,
+      "learning_rate": 1.3075250914142413e-08,
+      "loss": 0.7863,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9818076812012706,
+      "grad_norm": 0.22358852624893188,
+      "learning_rate": 8.956455928476049e-09,
+      "loss": 0.7965,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9856579074020598,
+      "grad_norm": 0.23945052921772003,
+      "learning_rate": 5.6139601453786275e-09,
+      "loss": 0.7874,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9895081336028492,
+      "grad_norm": 0.23508386313915253,
+      "learning_rate": 3.0482834514805602e-09,
+      "loss": 0.7759,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9933583598036385,
+      "grad_norm": 0.21108676493167877,
+      "learning_rate": 1.2598249081713366e-09,
+      "loss": 0.7714,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9972085860044277,
+      "grad_norm": 0.25175976753234863,
+      "learning_rate": 2.488626895302826e-10,
+      "loss": 0.8013,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9999037443449803,
+      "step": 2597,
+      "total_flos": 4.3217624287286395e+18,
+      "train_loss": 0.8216921990313436,
+      "train_runtime": 10886.4378,
+      "train_samples_per_second": 61.076,
+      "train_steps_per_second": 0.239
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2597,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.3217624287286395e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3/mc2_l14_224/finetune/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e99ab262737247aa77b67532397105fb504b37cf6976195b739885e868e950c
+size 6136

continuous/qwen3/mc2_l14_224/pretrain/config.json

@@ -0,0 +1,60 @@
+{
+  "_attn_implementation_autoset": true,
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "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,
+  "max_position_embeddings": 40960,
+  "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,
+  "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",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "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/qwen3/mc2_l14_224/pretrain/mm_projector.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24a6e7c16e33cd6afafa8c300df52f4c373122f553b82bae9d865e8a0133cd0c
+size 25184312

continuous/qwen3/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": 12.372112274169922,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 8.8013,
+      "step": 10
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 1.9714665412902832,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 5.5922,
+      "step": 20
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.13537639379501343,
+      "learning_rate": 0.0009999907688176173,
+      "loss": 4.871,
+      "step": 30
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.12968586385250092,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 4.7061,
+      "step": 40
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.10379946976900101,
+      "learning_rate": 0.0009959345572940772,
+      "loss": 4.6324,
+      "step": 50
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.09616631269454956,
+      "learning_rate": 0.0009911550081155983,
+      "loss": 4.5845,
+      "step": 60
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.12307918816804886,
+      "learning_rate": 0.0009845624347566433,
+      "loss": 4.5289,
+      "step": 70
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.12561112642288208,
+      "learning_rate": 0.0009761811727013547,
+      "loss": 4.4944,
+      "step": 80
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.09960111975669861,
+      "learning_rate": 0.0009660421601067666,
+      "loss": 4.445,
+      "step": 90
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.10494697839021683,
+      "learning_rate": 0.0009541828235992926,
+      "loss": 4.4121,
+      "step": 100
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.13056960701942444,
+      "learning_rate": 0.0009406469401200151,
+      "loss": 4.3755,
+      "step": 110
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.21604369580745697,
+      "learning_rate": 0.0009254844753287492,
+      "loss": 4.3557,
+      "step": 120
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.17312116920948029,
+      "learning_rate": 0.0009087513991633924,
+      "loss": 4.3204,
+      "step": 130
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.16003553569316864,
+      "learning_rate": 0.0008905094792353917,
+      "loss": 4.2972,
+      "step": 140
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.1496351659297943,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 4.2863,
+      "step": 150
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.14091233909130096,
+      "learning_rate": 0.0008497737783108237,
+      "loss": 4.2638,
+      "step": 160
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.14212998747825623,
+      "learning_rate": 0.0008274303669726426,
+      "loss": 4.2461,
+      "step": 170
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.13819803297519684,
+      "learning_rate": 0.0008038782961218137,
+      "loss": 4.2292,
+      "step": 180
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.13595464825630188,
+      "learning_rate": 0.000779204504653894,
+      "loss": 4.2216,
+      "step": 190
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.14357292652130127,
+      "learning_rate": 0.0007535000721258791,
+      "loss": 4.1997,
+      "step": 200
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.14223651587963104,
+      "learning_rate": 0.00072685988254984,
+      "loss": 4.1883,
+      "step": 210
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.12323244661092758,
+      "learning_rate": 0.0006993822741429906,
+      "loss": 4.201,
+      "step": 220
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.13006684184074402,
+      "learning_rate": 0.0006711686763270818,
+      "loss": 4.1943,
+      "step": 230
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.14380347728729248,
+      "learning_rate": 0.0006423232353170831,
+      "loss": 4.1782,
+      "step": 240
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.12392624467611313,
+      "learning_rate": 0.0006129524296812335,
+      "loss": 4.161,
+      "step": 250
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.11815076321363449,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 4.1726,
+      "step": 260
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.16648714244365692,
+      "learning_rate": 0.0005530699351157809,
+      "loss": 4.1508,
+      "step": 270
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.11089617013931274,
+      "learning_rate": 0.0005227792933277943,
+      "loss": 4.147,
+      "step": 280
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.13568514585494995,
+      "learning_rate": 0.0004924045652352092,
+      "loss": 4.1347,
+      "step": 290
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.15855349600315094,
+      "learning_rate": 0.0004620578745374633,
+      "loss": 4.1338,
+      "step": 300
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.13085976243019104,
+      "learning_rate": 0.0004318512414382059,
+      "loss": 4.1465,
+      "step": 310
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.14032915234565735,
+      "learning_rate": 0.0004018961691397148,
+      "loss": 4.1369,
+      "step": 320
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.11636923253536224,
+      "learning_rate": 0.0003723032322457458,
+      "loss": 4.1191,
+      "step": 330
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.17285268008708954,
+      "learning_rate": 0.0003431816685921625,
+      "loss": 4.1207,
+      "step": 340
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.14340779185295105,
+      "learning_rate": 0.0003146389760120416,
+      "loss": 4.121,
+      "step": 350
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.14484703540802002,
+      "learning_rate": 0.00028678051552373485,
+      "loss": 4.1282,
+      "step": 360
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.10880780220031738,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 4.1125,
+      "step": 370
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.13657166063785553,
+      "learning_rate": 0.0002335247266004629,
+      "loss": 4.1267,
+      "step": 380
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.11751321703195572,
+      "learning_rate": 0.000208323983828828,
+      "loss": 4.1062,
+      "step": 390
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.10982247442007065,
+      "learning_rate": 0.0001841999188095224,
+      "loss": 4.1118,
+      "step": 400
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.10838630050420761,
+      "learning_rate": 0.00016124158186777676,
+      "loss": 4.1082,
+      "step": 410
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.10520922392606735,
+      "learning_rate": 0.0001395337202205211,
+      "loss": 4.0996,
+      "step": 420
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.10388272255659103,
+      "learning_rate": 0.00011915646514489292,
+      "loss": 4.1065,
+      "step": 430
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.12091171741485596,
+      "learning_rate": 0.00010018503618578818,
+      "loss": 4.1119,
+      "step": 440
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.11123461276292801,
+      "learning_rate": 8.268946349432583e-05,
+      "loss": 4.1038,
+      "step": 450
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.11209961026906967,
+      "learning_rate": 6.673432932217117e-05,
+      "loss": 4.0953,
+      "step": 460
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.10920216888189316,
+      "learning_rate": 5.2378529625954684e-05,
+      "loss": 4.0842,
+      "step": 470
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.1166127622127533,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 4.0946,
+      "step": 480
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.10020587593317032,
+      "learning_rate": 2.8670803372379538e-05,
+      "loss": 4.0957,
+      "step": 490
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.10208002477884293,
+      "learning_rate": 1.9406390288876586e-05,
+      "loss": 4.1067,
+      "step": 500
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.10199034214019775,
+      "learning_rate": 1.191601558631461e-05,
+      "loss": 4.0963,
+      "step": 510
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.09626436233520508,
+      "learning_rate": 6.2273288462559305e-06,
+      "loss": 4.0893,
+      "step": 520
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.10138488560914993,
+      "learning_rate": 2.361328992552314e-06,
+      "loss": 4.0903,
+      "step": 530
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.10531555861234665,
+      "learning_rate": 3.322867770044202e-07,
+      "loss": 4.0972,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "step": 545,
+      "total_flos": 3.4196958409549414e+17,
+      "train_loss": 4.334117766914018,
+      "train_runtime": 5639.2018,
+      "train_samples_per_second": 98.973,
+      "train_steps_per_second": 0.097
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 545,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.4196958409549414e+17,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3/mc2_m16_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/qwen3/mc2_m16_224/finetune/config.json

@@ -0,0 +1,59 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "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,
+  "max_position_embeddings": 40960,
+  "max_window_layers": 28,
+  "mm_hidden_size": 512,
+  "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-M-16",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "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",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "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_m16_224.pt",
+  "vocab_size": 151936
+}

continuous/qwen3/mc2_m16_224/finetune/generation_config.json

@@ -0,0 +1,13 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.6,
+  "top_k": 20,
+  "top_p": 0.95,
+  "transformers_version": "4.51.3"
+}

continuous/qwen3/mc2_m16_224/finetune/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bed35a115940572a9da6bbbc45a48286e6166f50140ce54e7a5978704a86da2b
+size 4161700600

continuous/qwen3/mc2_m16_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/mc2_m16_224/finetune/tokenizer.json

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

continuous/qwen3/mc2_m16_224/finetune/tokenizer_config.json

@@ -0,0 +1,241 @@
+{
+  "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,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0].role == 'system' %}\n        {{- messages[0].content + '\\n\\n' }}\n    {%- endif %}\n    {{- \"# 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>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\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\" }}\n{%- else %}\n    {%- if messages[0].role == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0].content + '<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n    {%- set index = (messages|length - 1) - loop.index0 %}\n    {%- if ns.multi_step_tool and message.role == \"user\" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n        {%- set ns.multi_step_tool = false %}\n        {%- set ns.last_query_index = index %}\n    {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n    {%- if message.content is string %}\n        {%- set content = message.content %}\n    {%- else %}\n        {%- set content = '' %}\n    {%- endif %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set reasoning_content = '' %}\n        {%- if message.reasoning_content is string %}\n            {%- set reasoning_content = message.reasoning_content %}\n        {%- else %}\n            {%- if '</think>' in content %}\n                {%- set reasoning_content = content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n                {%- set content = content.split('</think>')[-1].lstrip('\\n') %}\n            {%- endif %}\n        {%- endif %}\n        {%- if loop.index0 > ns.last_query_index %}\n            {%- if loop.last or (not loop.last and reasoning_content) %}\n                {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n            {%- else %}\n                {{- '<|im_start|>' + message.role + '\\n' + content }}\n            {%- endif %}\n        {%- else %}\n            {{- '<|im_start|>' + message.role + '\\n' + content }}\n        {%- endif %}\n        {%- if message.tool_calls %}\n            {%- for tool_call in message.tool_calls %}\n                {%- if (loop.first and content) or (not loop.first) %}\n                    {{- '\\n' }}\n                {%- endif %}\n                {%- if tool_call.function %}\n                    {%- set tool_call = tool_call.function %}\n                {%- endif %}\n                {{- '<tool_call>\\n{\"name\": \"' }}\n                {{- tool_call.name }}\n                {{- '\", \"arguments\": ' }}\n                {%- if tool_call.arguments is string %}\n                    {{- tool_call.arguments }}\n                {%- else %}\n                    {{- tool_call.arguments | tojson }}\n                {%- endif %}\n                {{- '}\\n</tool_call>' }}\n            {%- endfor %}\n        {%- endif %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n    {%- if enable_thinking is defined and enable_thinking is false %}\n        {{- '<think>\\n\\n</think>\\n\\n' }}\n    {%- endif %}\n{%- endif %}",
+  "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/mc2_m16_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": 17.964662551879883,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 2.2898,
+      "step": 10
+    },
+    {
+      "epoch": 0.007700452401578592,
+      "grad_norm": 6.60088586807251,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.9065,
+      "step": 20
+    },
+    {
+      "epoch": 0.01155067860236789,
+      "grad_norm": 4.47412109375,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.5494,
+      "step": 30
+    },
+    {
+      "epoch": 0.015400904803157185,
+      "grad_norm": 1.1584409475326538,
+      "learning_rate": 5e-06,
+      "loss": 1.212,
+      "step": 40
+    },
+    {
+      "epoch": 0.019251131003946483,
+      "grad_norm": 0.7746429443359375,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.0901,
+      "step": 50
+    },
+    {
+      "epoch": 0.02310135720473578,
+      "grad_norm": 0.4157203733921051,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.998,
+      "step": 60
+    },
+    {
+      "epoch": 0.026951583405525073,
+      "grad_norm": 0.3320835828781128,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.9456,
+      "step": 70
+    },
+    {
+      "epoch": 0.03080180960631437,
+      "grad_norm": 0.30085793137550354,
+      "learning_rate": 9.999996111488723e-06,
+      "loss": 0.9434,
+      "step": 80
+    },
+    {
+      "epoch": 0.034652035807103666,
+      "grad_norm": 0.2820378541946411,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 0.9089,
+      "step": 90
+    },
+    {
+      "epoch": 0.038502262007892966,
+      "grad_norm": 0.2616160809993744,
+      "learning_rate": 9.998285264324193e-06,
+      "loss": 0.8907,
+      "step": 100
+    },
+    {
+      "epoch": 0.04235248820868226,
+      "grad_norm": 0.25084689259529114,
+      "learning_rate": 9.996263605626052e-06,
+      "loss": 0.871,
+      "step": 110
+    },
+    {
+      "epoch": 0.04620271440947156,
+      "grad_norm": 0.2508465051651001,
+      "learning_rate": 9.99346483580503e-06,
+      "loss": 0.8803,
+      "step": 120
+    },
+    {
+      "epoch": 0.05005294061026085,
+      "grad_norm": 0.2566666603088379,
+      "learning_rate": 9.989889390177467e-06,
+      "loss": 0.876,
+      "step": 130
+    },
+    {
+      "epoch": 0.053903166811050146,
+      "grad_norm": 0.27821531891822815,
+      "learning_rate": 9.985537824862647e-06,
+      "loss": 0.8693,
+      "step": 140
+    },
+    {
+      "epoch": 0.057753393011839446,
+      "grad_norm": 0.2955886721611023,
+      "learning_rate": 9.980410816696317e-06,
+      "loss": 0.8543,
+      "step": 150
+    },
+    {
+      "epoch": 0.06160361921262874,
+      "grad_norm": 0.25568777322769165,
+      "learning_rate": 9.97450916312541e-06,
+      "loss": 0.8588,
+      "step": 160
+    },
+    {
+      "epoch": 0.06545384541341803,
+      "grad_norm": 0.2242199033498764,
+      "learning_rate": 9.967833782084004e-06,
+      "loss": 0.8593,
+      "step": 170
+    },
+    {
+      "epoch": 0.06930407161420733,
+      "grad_norm": 0.2685142755508423,
+      "learning_rate": 9.960385711850554e-06,
+      "loss": 0.8868,
+      "step": 180
+    },
+    {
+      "epoch": 0.07315429781499663,
+      "grad_norm": 0.25718340277671814,
+      "learning_rate": 9.952166110886394e-06,
+      "loss": 0.8756,
+      "step": 190
+    },
+    {
+      "epoch": 0.07700452401578593,
+      "grad_norm": 0.2821897864341736,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.8653,
+      "step": 200
+    },
+    {
+      "epoch": 0.08085475021657522,
+      "grad_norm": 0.2523501217365265,
+      "learning_rate": 9.933417550425952e-06,
+      "loss": 0.8593,
+      "step": 210
+    },
+    {
+      "epoch": 0.08470497641736452,
+      "grad_norm": 0.256017804145813,
+      "learning_rate": 9.9228915070518e-06,
+      "loss": 0.847,
+      "step": 220
+    },
+    {
+      "epoch": 0.08855520261815382,
+      "grad_norm": 0.2521305978298187,
+      "learning_rate": 9.911599764737632e-06,
+      "loss": 0.8474,
+      "step": 230
+    },
+    {
+      "epoch": 0.09240542881894312,
+      "grad_norm": 0.2510559558868408,
+      "learning_rate": 9.899544079783606e-06,
+      "loss": 0.8422,
+      "step": 240
+    },
+    {
+      "epoch": 0.0962556550197324,
+      "grad_norm": 0.2893846333026886,
+      "learning_rate": 9.886726327312338e-06,
+      "loss": 0.8578,
+      "step": 250
+    },
+    {
+      "epoch": 0.1001058812205217,
+      "grad_norm": 0.24457940459251404,
+      "learning_rate": 9.87314850097724e-06,
+      "loss": 0.8547,
+      "step": 260
+    },
+    {
+      "epoch": 0.103956107421311,
+      "grad_norm": 0.25250476598739624,
+      "learning_rate": 9.858812712652452e-06,
+      "loss": 0.8422,
+      "step": 270
+    },
+    {
+      "epoch": 0.10780633362210029,
+      "grad_norm": 0.2378825694322586,
+      "learning_rate": 9.843721192104341e-06,
+      "loss": 0.8426,
+      "step": 280
+    },
+    {
+      "epoch": 0.11165655982288959,
+      "grad_norm": 0.24740518629550934,
+      "learning_rate": 9.827876286644698e-06,
+      "loss": 0.8589,
+      "step": 290
+    },
+    {
+      "epoch": 0.11550678602367889,
+      "grad_norm": 0.2306535691022873,
+      "learning_rate": 9.811280460765643e-06,
+      "loss": 0.8205,
+      "step": 300
+    },
+    {
+      "epoch": 0.11935701222446819,
+      "grad_norm": 0.24680525064468384,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.8332,
+      "step": 310
+    },
+    {
+      "epoch": 0.12320723842525748,
+      "grad_norm": 0.24220439791679382,
+      "learning_rate": 9.775846489301279e-06,
+      "loss": 0.8329,
+      "step": 320
+    },
+    {
+      "epoch": 0.12705746462604678,
+      "grad_norm": 0.24907219409942627,
+      "learning_rate": 9.757013855061157e-06,
+      "loss": 0.8328,
+      "step": 330
+    },
+    {
+      "epoch": 0.13090769082683607,
+      "grad_norm": 0.240876704454422,
+      "learning_rate": 9.737441322234758e-06,
+      "loss": 0.8442,
+      "step": 340
+    },
+    {
+      "epoch": 0.13475791702762538,
+      "grad_norm": 0.2558647096157074,
+      "learning_rate": 9.717131935103605e-06,
+      "loss": 0.8436,
+      "step": 350
+    },
+    {
+      "epoch": 0.13860814322841467,
+      "grad_norm": 0.24324782192707062,
+      "learning_rate": 9.696088852558398e-06,
+      "loss": 0.8302,
+      "step": 360
+    },
+    {
+      "epoch": 0.14245836942920398,
+      "grad_norm": 0.24751833081245422,
+      "learning_rate": 9.674315347607686e-06,
+      "loss": 0.8558,
+      "step": 370
+    },
+    {
+      "epoch": 0.14630859562999327,
+      "grad_norm": 0.2374933958053589,
+      "learning_rate": 9.651814806868796e-06,
+      "loss": 0.8164,
+      "step": 380
+    },
+    {
+      "epoch": 0.15015882183078255,
+      "grad_norm": 0.23850394785404205,
+      "learning_rate": 9.62859073004107e-06,
+      "loss": 0.8246,
+      "step": 390
+    },
+    {
+      "epoch": 0.15400904803157187,
+      "grad_norm": 0.23875148594379425,
+      "learning_rate": 9.604646729361547e-06,
+      "loss": 0.8272,
+      "step": 400
+    },
+    {
+      "epoch": 0.15785927423236115,
+      "grad_norm": 0.22836053371429443,
+      "learning_rate": 9.579986529043099e-06,
+      "loss": 0.8152,
+      "step": 410
+    },
+    {
+      "epoch": 0.16170950043315044,
+      "grad_norm": 0.24262051284313202,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.832,
+      "step": 420
+    },
+    {
+      "epoch": 0.16555972663393975,
+      "grad_norm": 0.2517402470111847,
+      "learning_rate": 9.528532982727281e-06,
+      "loss": 0.821,
+      "step": 430
+    },
+    {
+      "epoch": 0.16940995283472904,
+      "grad_norm": 0.24081404507160187,
+      "learning_rate": 9.501747639735018e-06,
+      "loss": 0.8186,
+      "step": 440
+    },
+    {
+      "epoch": 0.17326017903551832,
+      "grad_norm": 0.24995779991149902,
+      "learning_rate": 9.474262101869272e-06,
+      "loss": 0.8082,
+      "step": 450
+    },
+    {
+      "epoch": 0.17711040523630764,
+      "grad_norm": 0.238996684551239,
+      "learning_rate": 9.446080644188141e-06,
+      "loss": 0.813,
+      "step": 460
+    },
+    {
+      "epoch": 0.18096063143709692,
+      "grad_norm": 0.24162709712982178,
+      "learning_rate": 9.417207649992017e-06,
+      "loss": 0.8221,
+      "step": 470
+    },
+    {
+      "epoch": 0.18481085763788624,
+      "grad_norm": 0.24608567357063293,
+      "learning_rate": 9.387647610141816e-06,
+      "loss": 0.8202,
+      "step": 480
+    },
+    {
+      "epoch": 0.18866108383867553,
+      "grad_norm": 0.23444557189941406,
+      "learning_rate": 9.35740512236047e-06,
+      "loss": 0.8067,
+      "step": 490
+    },
+    {
+      "epoch": 0.1925113100394648,
+      "grad_norm": 0.2296661138534546,
+      "learning_rate": 9.326484890517809e-06,
+      "loss": 0.8113,
+      "step": 500
+    },
+    {
+      "epoch": 0.19636153624025413,
+      "grad_norm": 0.23526917397975922,
+      "learning_rate": 9.294891723898926e-06,
+      "loss": 0.8069,
+      "step": 510
+    },
+    {
+      "epoch": 0.2002117624410434,
+      "grad_norm": 0.2401217520236969,
+      "learning_rate": 9.262630536456153e-06,
+      "loss": 0.8372,
+      "step": 520
+    },
+    {
+      "epoch": 0.2040619886418327,
+      "grad_norm": 0.2527241110801697,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.808,
+      "step": 530
+    },
+    {
+      "epoch": 0.207912214842622,
+      "grad_norm": 0.2570526897907257,
+      "learning_rate": 9.196124273642422e-06,
+      "loss": 0.8217,
+      "step": 540
+    },
+    {
+      "epoch": 0.2117624410434113,
+      "grad_norm": 0.24740824103355408,
+      "learning_rate": 9.161889542552842e-06,
+      "loss": 0.8237,
+      "step": 550
+    },
+    {
+      "epoch": 0.21561266724420058,
+      "grad_norm": 0.23169489204883575,
+      "learning_rate": 9.127007477593193e-06,
+      "loss": 0.8281,
+      "step": 560
+    },
+    {
+      "epoch": 0.2194628934449899,
+      "grad_norm": 0.269357830286026,
+      "learning_rate": 9.091483504265974e-06,
+      "loss": 0.8374,
+      "step": 570
+    },
+    {
+      "epoch": 0.22331311964577918,
+      "grad_norm": 0.25544172525405884,
+      "learning_rate": 9.055323147915116e-06,
+      "loss": 0.8171,
+      "step": 580
+    },
+    {
+      "epoch": 0.2271633458465685,
+      "grad_norm": 0.25473591685295105,
+      "learning_rate": 9.018532032866585e-06,
+      "loss": 0.8278,
+      "step": 590
+    },
+    {
+      "epoch": 0.23101357204735778,
+      "grad_norm": 0.23406583070755005,
+      "learning_rate": 8.981115881553579e-06,
+      "loss": 0.8177,
+      "step": 600
+    },
+    {
+      "epoch": 0.23486379824814707,
+      "grad_norm": 0.2666371464729309,
+      "learning_rate": 8.943080513626476e-06,
+      "loss": 0.8376,
+      "step": 610
+    },
+    {
+      "epoch": 0.23871402444893638,
+      "grad_norm": 0.23917105793952942,
+      "learning_rate": 8.904431845047641e-06,
+      "loss": 0.8215,
+      "step": 620
+    },
+    {
+      "epoch": 0.24256425064972567,
+      "grad_norm": 0.254898339509964,
+      "learning_rate": 8.86517588717128e-06,
+      "loss": 0.8384,
+      "step": 630
+    },
+    {
+      "epoch": 0.24641447685051496,
+      "grad_norm": 0.2350817322731018,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.8267,
+      "step": 640
+    },
+    {
+      "epoch": 0.25026470305130427,
+      "grad_norm": 0.22207516431808472,
+      "learning_rate": 8.784866620277314e-06,
+      "loss": 0.7849,
+      "step": 650
+    },
+    {
+      "epoch": 0.25411492925209356,
+      "grad_norm": 0.25214889645576477,
+      "learning_rate": 8.743825802439018e-06,
+      "loss": 0.8213,
+      "step": 660
+    },
+    {
+      "epoch": 0.25796515545288284,
+      "grad_norm": 0.2673768401145935,
+      "learning_rate": 8.702202675718956e-06,
+      "loss": 0.8259,
+      "step": 670
+    },
+    {
+      "epoch": 0.26181538165367213,
+      "grad_norm": 0.24445728957653046,
+      "learning_rate": 8.660003714113959e-06,
+      "loss": 0.7959,
+      "step": 680
+    },
+    {
+      "epoch": 0.26566560785446147,
+      "grad_norm": 0.2406833916902542,
+      "learning_rate": 8.617235481185323e-06,
+      "loss": 0.8218,
+      "step": 690
+    },
+    {
+      "epoch": 0.26951583405525076,
+      "grad_norm": 0.2449120432138443,
+      "learning_rate": 8.573904629037931e-06,
+      "loss": 0.8265,
+      "step": 700
+    },
+    {
+      "epoch": 0.27336606025604004,
+      "grad_norm": 0.2410110980272293,
+      "learning_rate": 8.530017897285585e-06,
+      "loss": 0.8066,
+      "step": 710
+    },
+    {
+      "epoch": 0.27721628645682933,
+      "grad_norm": 0.22213158011436462,
+      "learning_rate": 8.485582112002746e-06,
+      "loss": 0.8096,
+      "step": 720
+    },
+    {
+      "epoch": 0.2810665126576186,
+      "grad_norm": 0.2446628361940384,
+      "learning_rate": 8.440604184662811e-06,
+      "loss": 0.8161,
+      "step": 730
+    },
+    {
+      "epoch": 0.28491673885840796,
+      "grad_norm": 0.22914154827594757,
+      "learning_rate": 8.395091111063118e-06,
+      "loss": 0.8042,
+      "step": 740
+    },
+    {
+      "epoch": 0.28876696505919724,
+      "grad_norm": 0.2251976728439331,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.8037,
+      "step": 750
+    },
+    {
+      "epoch": 0.29261719125998653,
+      "grad_norm": 0.24627074599266052,
+      "learning_rate": 8.302487923351842e-06,
+      "loss": 0.7963,
+      "step": 760
+    },
+    {
+      "epoch": 0.2964674174607758,
+      "grad_norm": 0.24244199693202972,
+      "learning_rate": 8.255412212597021e-06,
+      "loss": 0.8142,
+      "step": 770
+    },
+    {
+      "epoch": 0.3003176436615651,
+      "grad_norm": 0.23052042722702026,
+      "learning_rate": 8.207830160055688e-06,
+      "loss": 0.8104,
+      "step": 780
+    },
+    {
+      "epoch": 0.3041678698623544,
+      "grad_norm": 0.24605132639408112,
+      "learning_rate": 8.159749166566786e-06,
+      "loss": 0.7948,
+      "step": 790
+    },
+    {
+      "epoch": 0.30801809606314373,
+      "grad_norm": 0.23195461928844452,
+      "learning_rate": 8.11117671057377e-06,
+      "loss": 0.8227,
+      "step": 800
+    },
+    {
+      "epoch": 0.311868322263933,
+      "grad_norm": 0.24305589497089386,
+      "learning_rate": 8.062120346961404e-06,
+      "loss": 0.8156,
+      "step": 810
+    },
+    {
+      "epoch": 0.3157185484647223,
+      "grad_norm": 0.23656965792179108,
+      "learning_rate": 8.012587705880705e-06,
+      "loss": 0.8076,
+      "step": 820
+    },
+    {
+      "epoch": 0.3195687746655116,
+      "grad_norm": 0.24814729392528534,
+      "learning_rate": 7.962586491562146e-06,
+      "loss": 0.794,
+      "step": 830
+    },
+    {
+      "epoch": 0.3234190008663009,
+      "grad_norm": 0.2534838914871216,
+      "learning_rate": 7.912124481117357e-06,
+      "loss": 0.7996,
+      "step": 840
+    },
+    {
+      "epoch": 0.3272692270670902,
+      "grad_norm": 0.24142204225063324,
+      "learning_rate": 7.861209523329494e-06,
+      "loss": 0.8103,
+      "step": 850
+    },
+    {
+      "epoch": 0.3311194532678795,
+      "grad_norm": 0.2387242466211319,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.794,
+      "step": 860
+    },
+    {
+      "epoch": 0.3349696794686688,
+      "grad_norm": 0.24049800634384155,
+      "learning_rate": 7.758052511879035e-06,
+      "loss": 0.8035,
+      "step": 870
+    },
+    {
+      "epoch": 0.3388199056694581,
+      "grad_norm": 0.22327958047389984,
+      "learning_rate": 7.705826503098642e-06,
+      "loss": 0.7987,
+      "step": 880
+    },
+    {
+      "epoch": 0.34267013187024736,
+      "grad_norm": 0.23305052518844604,
+      "learning_rate": 7.653179634243979e-06,
+      "loss": 0.7988,
+      "step": 890
+    },
+    {
+      "epoch": 0.34652035807103665,
+      "grad_norm": 0.22430409491062164,
+      "learning_rate": 7.600120093927696e-06,
+      "loss": 0.8044,
+      "step": 900
+    },
+    {
+      "epoch": 0.350370584271826,
+      "grad_norm": 0.24581289291381836,
+      "learning_rate": 7.546656134948729e-06,
+      "loss": 0.7941,
+      "step": 910
+    },
+    {
+      "epoch": 0.3542208104726153,
+      "grad_norm": 0.2432182878255844,
+      "learning_rate": 7.492796073008661e-06,
+      "loss": 0.7922,
+      "step": 920
+    },
+    {
+      "epoch": 0.35807103667340456,
+      "grad_norm": 0.2603885233402252,
+      "learning_rate": 7.438548285418326e-06,
+      "loss": 0.7776,
+      "step": 930
+    },
+    {
+      "epoch": 0.36192126287419385,
+      "grad_norm": 0.25645849108695984,
+      "learning_rate": 7.3839212097948e-06,
+      "loss": 0.812,
+      "step": 940
+    },
+    {
+      "epoch": 0.36577148907498314,
+      "grad_norm": 0.24397337436676025,
+      "learning_rate": 7.328923342749039e-06,
+      "loss": 0.827,
+      "step": 950
+    },
+    {
+      "epoch": 0.3696217152757725,
+      "grad_norm": 0.25670453906059265,
+      "learning_rate": 7.273563238564319e-06,
+      "loss": 0.8101,
+      "step": 960
+    },
+    {
+      "epoch": 0.37347194147656176,
+      "grad_norm": 0.24130189418792725,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.791,
+      "step": 970
+    },
+    {
+      "epoch": 0.37732216767735105,
+      "grad_norm": 0.23203547298908234,
+      "learning_rate": 7.161790816280858e-06,
+      "loss": 0.796,
+      "step": 980
+    },
+    {
+      "epoch": 0.38117239387814034,
+      "grad_norm": 0.24485979974269867,
+      "learning_rate": 7.1053958830920125e-06,
+      "loss": 0.7823,
+      "step": 990
+    },
+    {
+      "epoch": 0.3850226200789296,
+      "grad_norm": 0.23367318511009216,
+      "learning_rate": 7.0486734798799715e-06,
+      "loss": 0.7982,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3888728462797189,
+      "grad_norm": 0.2396024614572525,
+      "learning_rate": 6.991632429159708e-06,
+      "loss": 0.8212,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39272307248050825,
+      "grad_norm": 0.25694042444229126,
+      "learning_rate": 6.934281603008132e-06,
+      "loss": 0.812,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39657329868129754,
+      "grad_norm": 0.24822570383548737,
+      "learning_rate": 6.876629921684143e-06,
+      "loss": 0.8086,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4004235248820868,
+      "grad_norm": 0.24965888261795044,
+      "learning_rate": 6.818686352241199e-06,
+      "loss": 0.7992,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4042737510828761,
+      "grad_norm": 0.22224882245063782,
+      "learning_rate": 6.760459907132573e-06,
+      "loss": 0.7951,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4081239772836654,
+      "grad_norm": 0.22797204554080963,
+      "learning_rate": 6.701959642809589e-06,
+      "loss": 0.8011,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41197420348445474,
+      "grad_norm": 0.25105637311935425,
+      "learning_rate": 6.643194658312987e-06,
+      "loss": 0.7938,
+      "step": 1070
+    },
+    {
+      "epoch": 0.415824429685244,
+      "grad_norm": 0.2371562123298645,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.8163,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4196746558860333,
+      "grad_norm": 0.27385076880455017,
+      "learning_rate": 6.5249071294110735e-06,
+      "loss": 0.7958,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4235248820868226,
+      "grad_norm": 0.2536002993583679,
+      "learning_rate": 6.465402983265274e-06,
+      "loss": 0.8108,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4273751082876119,
+      "grad_norm": 0.25096264481544495,
+      "learning_rate": 6.405670910603243e-06,
+      "loss": 0.8049,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43122533448840117,
+      "grad_norm": 0.2637416422367096,
+      "learning_rate": 6.345720202059291e-06,
+      "loss": 0.8006,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4350755606891905,
+      "grad_norm": 0.25268760323524475,
+      "learning_rate": 6.285560182274008e-06,
+      "loss": 0.7794,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4389257868899798,
+      "grad_norm": 0.2352854460477829,
+      "learning_rate": 6.225200208443936e-06,
+      "loss": 0.8007,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4427760130907691,
+      "grad_norm": 0.23188091814517975,
+      "learning_rate": 6.16464966886616e-06,
+      "loss": 0.7901,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44662623929155837,
+      "grad_norm": 0.2812661826610565,
+      "learning_rate": 6.103917981478065e-06,
+      "loss": 0.8131,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45047646549234766,
+      "grad_norm": 0.24056114256381989,
+      "learning_rate": 6.043014592392489e-06,
+      "loss": 0.8237,
+      "step": 1170
+    },
+    {
+      "epoch": 0.454326691693137,
+      "grad_norm": 0.23404760658740997,
+      "learning_rate": 5.981948974428485e-06,
+      "loss": 0.7988,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4581769178939263,
+      "grad_norm": 0.2664545178413391,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.7767,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46202714409471557,
+      "grad_norm": 0.24890075623989105,
+      "learning_rate": 5.8593690678282355e-06,
+      "loss": 0.7876,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46587737029550486,
+      "grad_norm": 0.25923803448677063,
+      "learning_rate": 5.797873845081301e-06,
+      "loss": 0.7908,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46972759649629414,
+      "grad_norm": 0.24868686497211456,
+      "learning_rate": 5.736254522269075e-06,
+      "loss": 0.8068,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47357782269708343,
+      "grad_norm": 0.23769183456897736,
+      "learning_rate": 5.6745206835658386e-06,
+      "loss": 0.8093,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47742804889787277,
+      "grad_norm": 0.23292513191699982,
+      "learning_rate": 5.612681930957495e-06,
+      "loss": 0.7806,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48127827509866206,
+      "grad_norm": 0.23023663461208344,
+      "learning_rate": 5.550747882748095e-06,
+      "loss": 0.7776,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48512850129945134,
+      "grad_norm": 0.26334813237190247,
+      "learning_rate": 5.4887281720638195e-06,
+      "loss": 0.7837,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48897872750024063,
+      "grad_norm": 0.28428661823272705,
+      "learning_rate": 5.42663244535466e-06,
+      "loss": 0.7692,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4928289537010299,
+      "grad_norm": 0.2659589350223541,
+      "learning_rate": 5.3644703608940205e-06,
+      "loss": 0.7765,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49667917990181926,
+      "grad_norm": 0.24645563960075378,
+      "learning_rate": 5.302251587276491e-06,
+      "loss": 0.7882,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5005294061026085,
+      "grad_norm": 0.22712987661361694,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.787,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5043796323033978,
+      "grad_norm": 0.23621056973934174,
+      "learning_rate": 5.177682689530615e-06,
+      "loss": 0.8026,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5082298585041871,
+      "grad_norm": 0.2508206367492676,
+      "learning_rate": 5.11535194065619e-06,
+      "loss": 0.7697,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5120800847049765,
+      "grad_norm": 0.267014741897583,
+      "learning_rate": 5.053003250119113e-06,
+      "loss": 0.7796,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5159303109057657,
+      "grad_norm": 0.24479186534881592,
+      "learning_rate": 4.9906463155383925e-06,
+      "loss": 0.7873,
+      "step": 1340
+    },
+    {
+      "epoch": 0.519780537106555,
+      "grad_norm": 0.24559848010540009,
+      "learning_rate": 4.9282908358153046e-06,
+      "loss": 0.7795,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5236307633073443,
+      "grad_norm": 0.26693612337112427,
+      "learning_rate": 4.8659465096248345e-06,
+      "loss": 0.7967,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5274809895081336,
+      "grad_norm": 0.26651403307914734,
+      "learning_rate": 4.803623033907169e-06,
+      "loss": 0.7951,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5313312157089229,
+      "grad_norm": 0.23618242144584656,
+      "learning_rate": 4.741330102359441e-06,
+      "loss": 0.8006,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5351814419097122,
+      "grad_norm": 0.26320376992225647,
+      "learning_rate": 4.679077403927996e-06,
+      "loss": 0.7808,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5390316681105015,
+      "grad_norm": 0.24006524682044983,
+      "learning_rate": 4.616874621301373e-06,
+      "loss": 0.7802,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5428818943112907,
+      "grad_norm": 0.24921643733978271,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.7953,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5467321205120801,
+      "grad_norm": 0.2650436460971832,
+      "learning_rate": 4.492657493892822e-06,
+      "loss": 0.779,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5505823467128694,
+      "grad_norm": 0.2740040123462677,
+      "learning_rate": 4.430662469650999e-06,
+      "loss": 0.7942,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5544325729136587,
+      "grad_norm": 0.26734471321105957,
+      "learning_rate": 4.368755999289118e-06,
+      "loss": 0.7971,
+      "step": 1440
+    },
+    {
+      "epoch": 0.558282799114448,
+      "grad_norm": 0.27225741744041443,
+      "learning_rate": 4.306947711643949e-06,
+      "loss": 0.7817,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5621330253152372,
+      "grad_norm": 0.25284478068351746,
+      "learning_rate": 4.245247220281068e-06,
+      "loss": 0.7849,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5659832515160266,
+      "grad_norm": 0.2419055998325348,
+      "learning_rate": 4.183664121999587e-06,
+      "loss": 0.7817,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5698334777168159,
+      "grad_norm": 0.25292760133743286,
+      "learning_rate": 4.122207995339474e-06,
+      "loss": 0.7788,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5736837039176051,
+      "grad_norm": 0.23851096630096436,
+      "learning_rate": 4.060888399091734e-06,
+      "loss": 0.7837,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5775339301183945,
+      "grad_norm": 0.2318463921546936,
+      "learning_rate": 3.999714870811639e-06,
+      "loss": 0.7925,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5813841563191837,
+      "grad_norm": 0.23393011093139648,
+      "learning_rate": 3.938696925335275e-06,
+      "loss": 0.7851,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5852343825199731,
+      "grad_norm": 0.24212098121643066,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.793,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5890846087207623,
+      "grad_norm": 0.24901717901229858,
+      "learning_rate": 3.817165719666382e-06,
+      "loss": 0.7945,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5929348349215516,
+      "grad_norm": 0.2435370832681656,
+      "learning_rate": 3.756671362249841e-06,
+      "loss": 0.7993,
+      "step": 1540
+    },
+    {
+      "epoch": 0.596785061122341,
+      "grad_norm": 0.2412971407175064,
+      "learning_rate": 3.6963703902488907e-06,
+      "loss": 0.7846,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6006352873231302,
+      "grad_norm": 0.24110092222690582,
+      "learning_rate": 3.6362721827835655e-06,
+      "loss": 0.8021,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6044855135239195,
+      "grad_norm": 0.26334530115127563,
+      "learning_rate": 3.576386087436212e-06,
+      "loss": 0.7975,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6083357397247088,
+      "grad_norm": 0.253802627325058,
+      "learning_rate": 3.5167214187975913e-06,
+      "loss": 0.7871,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6121859659254981,
+      "grad_norm": 0.21041545271873474,
+      "learning_rate": 3.457287457018086e-06,
+      "loss": 0.7899,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6160361921262875,
+      "grad_norm": 0.23896731436252594,
+      "learning_rate": 3.3980934463643003e-06,
+      "loss": 0.7771,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6198864183270767,
+      "grad_norm": 0.22350122034549713,
+      "learning_rate": 3.3391485937812053e-06,
+      "loss": 0.7604,
+      "step": 1610
+    },
+    {
+      "epoch": 0.623736644527866,
+      "grad_norm": 0.2404075264930725,
+      "learning_rate": 3.280462067460115e-06,
+      "loss": 0.7581,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6275868707286553,
+      "grad_norm": 0.2488982230424881,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.7887,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6314370969294446,
+      "grad_norm": 0.21580159664154053,
+      "learning_rate": 3.163900464051089e-06,
+      "loss": 0.7798,
+      "step": 1640
+    },
+    {
+      "epoch": 0.635287323130234,
+      "grad_norm": 0.2441449761390686,
+      "learning_rate": 3.1060435167748826e-06,
+      "loss": 0.7889,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6391375493310232,
+      "grad_norm": 0.24121089279651642,
+      "learning_rate": 3.048481152564253e-06,
+      "loss": 0.7693,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6429877755318125,
+      "grad_norm": 0.2503420114517212,
+      "learning_rate": 2.9912223245804017e-06,
+      "loss": 0.792,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6468380017326018,
+      "grad_norm": 0.2500740587711334,
+      "learning_rate": 2.934275938772981e-06,
+      "loss": 0.787,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6506882279333911,
+      "grad_norm": 0.25226929783821106,
+      "learning_rate": 2.8776508524948676e-06,
+      "loss": 0.7767,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6545384541341804,
+      "grad_norm": 0.2178998440504074,
+      "learning_rate": 2.821355873124503e-06,
+      "loss": 0.7745,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6583886803349697,
+      "grad_norm": 0.25741666555404663,
+      "learning_rate": 2.7653997566960144e-06,
+      "loss": 0.7905,
+      "step": 1710
+    },
+    {
+      "epoch": 0.662238906535759,
+      "grad_norm": 0.2572144567966461,
+      "learning_rate": 2.709791206537309e-06,
+      "loss": 0.7794,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6660891327365482,
+      "grad_norm": 0.24554477632045746,
+      "learning_rate": 2.6545388719163747e-06,
+      "loss": 0.7849,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6699393589373376,
+      "grad_norm": 0.23566469550132751,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.7836,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6737895851381268,
+      "grad_norm": 0.2500940263271332,
+      "learning_rate": 2.545137167997007e-06,
+      "loss": 0.7794,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6776398113389162,
+      "grad_norm": 0.2279323786497116,
+      "learning_rate": 2.4910048148705972e-06,
+      "loss": 0.7672,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6814900375397055,
+      "grad_norm": 0.25420552492141724,
+      "learning_rate": 2.4372627069793327e-06,
+      "loss": 0.8111,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6853402637404947,
+      "grad_norm": 0.24456627666950226,
+      "learning_rate": 2.3839192032876486e-06,
+      "loss": 0.7965,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6891904899412841,
+      "grad_norm": 0.23925970494747162,
+      "learning_rate": 2.3309826007617047e-06,
+      "loss": 0.8082,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6930407161420733,
+      "grad_norm": 0.23509013652801514,
+      "learning_rate": 2.278461133078881e-06,
+      "loss": 0.7716,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6968909423428626,
+      "grad_norm": 0.24481487274169922,
+      "learning_rate": 2.2263629693471258e-06,
+      "loss": 0.7623,
+      "step": 1810
+    },
+    {
+      "epoch": 0.700741168543652,
+      "grad_norm": 0.23325252532958984,
+      "learning_rate": 2.1746962128343414e-06,
+      "loss": 0.7867,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7045913947444412,
+      "grad_norm": 0.23235656321048737,
+      "learning_rate": 2.1234688997080183e-06,
+      "loss": 0.7753,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7084416209452306,
+      "grad_norm": 0.2552795708179474,
+      "learning_rate": 2.0726889977853053e-06,
+      "loss": 0.7803,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7122918471460198,
+      "grad_norm": 0.2528349459171295,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.794,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7161420733468091,
+      "grad_norm": 0.23485736548900604,
+      "learning_rate": 1.9725029496425814e-06,
+      "loss": 0.7858,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7199922995475985,
+      "grad_norm": 0.2491462379693985,
+      "learning_rate": 1.923112386205727e-06,
+      "loss": 0.783,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7238425257483877,
+      "grad_norm": 0.22504505515098572,
+      "learning_rate": 1.874200397115079e-06,
+      "loss": 0.7899,
+      "step": 1880
+    },
+    {
+      "epoch": 0.727692751949177,
+      "grad_norm": 0.2522429823875427,
+      "learning_rate": 1.825774590065857e-06,
+      "loss": 0.7747,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7315429781499663,
+      "grad_norm": 0.25222674012184143,
+      "learning_rate": 1.7778424971332813e-06,
+      "loss": 0.7805,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7353932043507556,
+      "grad_norm": 0.24080727994441986,
+      "learning_rate": 1.7304115736010407e-06,
+      "loss": 0.779,
+      "step": 1910
+    },
+    {
+      "epoch": 0.739243430551545,
+      "grad_norm": 0.2409914880990982,
+      "learning_rate": 1.6834891968017108e-06,
+      "loss": 0.7726,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7430936567523342,
+      "grad_norm": 0.2459801435470581,
+      "learning_rate": 1.6370826649692884e-06,
+      "loss": 0.7907,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7469438829531235,
+      "grad_norm": 0.2434801310300827,
+      "learning_rate": 1.5911991961040463e-06,
+      "loss": 0.7981,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7507941091539128,
+      "grad_norm": 0.24736011028289795,
+      "learning_rate": 1.545845926849851e-06,
+      "loss": 0.7858,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7546443353547021,
+      "grad_norm": 0.2562319338321686,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.8017,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7584945615554913,
+      "grad_norm": 0.24396784603595734,
+      "learning_rate": 1.4567581203207238e-06,
+      "loss": 0.7925,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7623447877562807,
+      "grad_norm": 0.25736188888549805,
+      "learning_rate": 1.413037439625593e-06,
+      "loss": 0.7892,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76619501395707,
+      "grad_norm": 0.26442015171051025,
+      "learning_rate": 1.3698746695458847e-06,
+      "loss": 0.7854,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7700452401578592,
+      "grad_norm": 0.23835445940494537,
+      "learning_rate": 1.3272765235521745e-06,
+      "loss": 0.7775,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7738954663586486,
+      "grad_norm": 0.23767007887363434,
+      "learning_rate": 1.2852496272942877e-06,
+      "loss": 0.7841,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7777456925594378,
+      "grad_norm": 0.23074036836624146,
+      "learning_rate": 1.2438005175707396e-06,
+      "loss": 0.7728,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7815959187602272,
+      "grad_norm": 0.23877929151058197,
+      "learning_rate": 1.2029356413120297e-06,
+      "loss": 0.757,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7854461449610165,
+      "grad_norm": 0.2509441077709198,
+      "learning_rate": 1.1626613545778786e-06,
+      "loss": 0.775,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7892963711618057,
+      "grad_norm": 0.2301674485206604,
+      "learning_rate": 1.1229839215686294e-06,
+      "loss": 0.7792,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7931465973625951,
+      "grad_norm": 0.219220831990242,
+      "learning_rate": 1.0839095136509188e-06,
+      "loss": 0.7633,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7969968235633843,
+      "grad_norm": 0.25800326466560364,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.786,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8008470497641736,
+      "grad_norm": 0.23034441471099854,
+      "learning_rate": 1.007593988643402e-06,
+      "loss": 0.7668,
+      "step": 2080
+    },
+    {
+      "epoch": 0.804697275964963,
+      "grad_norm": 0.2453928142786026,
+      "learning_rate": 9.703647415524613e-07,
+      "loss": 0.7884,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8085475021657522,
+      "grad_norm": 0.23805692791938782,
+      "learning_rate": 9.337622577045485e-07,
+      "loss": 0.7773,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8123977283665416,
+      "grad_norm": 0.2369479238986969,
+      "learning_rate": 8.977922301934588e-07,
+      "loss": 0.7869,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8162479545673308,
+      "grad_norm": 0.23941001296043396,
+      "learning_rate": 8.624602537417009e-07,
+      "loss": 0.7887,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8200981807681201,
+      "grad_norm": 0.26144489645957947,
+      "learning_rate": 8.277718238303123e-07,
+      "loss": 0.7764,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8239484069689095,
+      "grad_norm": 0.23836737871170044,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.7794,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8277986331696987,
+      "grad_norm": 0.24814127385616302,
+      "learning_rate": 7.603470842324223e-07,
+      "loss": 0.7724,
+      "step": 2150
+    },
+    {
+      "epoch": 0.831648859370488,
+      "grad_norm": 0.24859094619750977,
+      "learning_rate": 7.276212616857575e-07,
+      "loss": 0.7884,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8354990855712773,
+      "grad_norm": 0.2296704202890396,
+      "learning_rate": 6.955599583279837e-07,
+      "loss": 0.7882,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8393493117720666,
+      "grad_norm": 0.24677014350891113,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.7801,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8431995379728559,
+      "grad_norm": 0.23567335307598114,
+      "learning_rate": 6.334507521075711e-07,
+      "loss": 0.7868,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8470497641736452,
+      "grad_norm": 0.25627148151397705,
+      "learning_rate": 6.034125096149179e-07,
+      "loss": 0.7811,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8508999903744345,
+      "grad_norm": 0.2391371726989746,
+      "learning_rate": 5.740581055485722e-07,
+      "loss": 0.7537,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8547502165752238,
+      "grad_norm": 0.22977998852729797,
+      "learning_rate": 5.453921056471961e-07,
+      "loss": 0.7874,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8586004427760131,
+      "grad_norm": 0.24648508429527283,
+      "learning_rate": 5.174189685761288e-07,
+      "loss": 0.7808,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8624506689768023,
+      "grad_norm": 0.23217733204364777,
+      "learning_rate": 4.901430452338979e-07,
+      "loss": 0.7927,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8663008951775917,
+      "grad_norm": 0.24390576779842377,
+      "learning_rate": 4.6356857807548286e-07,
+      "loss": 0.7756,
+      "step": 2250
+    },
+    {
+      "epoch": 0.870151121378381,
+      "grad_norm": 0.24087175726890564,
+      "learning_rate": 4.376997004524541e-07,
+      "loss": 0.7644,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8740013475791703,
+      "grad_norm": 0.24330322444438934,
+      "learning_rate": 4.12540435970073e-07,
+      "loss": 0.7744,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8778515737799596,
+      "grad_norm": 0.23751062154769897,
+      "learning_rate": 3.880946978614703e-07,
+      "loss": 0.7866,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8817017999807488,
+      "grad_norm": 0.24107448756694794,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.7896,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8855520261815382,
+      "grad_norm": 0.22517569363117218,
+      "learning_rate": 3.41358898202781e-07,
+      "loss": 0.767,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8894022523823275,
+      "grad_norm": 0.22279219329357147,
+      "learning_rate": 3.190761058667741e-07,
+      "loss": 0.7782,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8932524785831167,
+      "grad_norm": 0.23094573616981506,
+      "learning_rate": 2.9752137720206463e-07,
+      "loss": 0.784,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8971027047839061,
+      "grad_norm": 0.241532564163208,
+      "learning_rate": 2.7669806479785067e-07,
+      "loss": 0.7886,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9009529309846953,
+      "grad_norm": 0.2454252690076828,
+      "learning_rate": 2.566094074799763e-07,
+      "loss": 0.7831,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9048031571854847,
+      "grad_norm": 0.23480668663978577,
+      "learning_rate": 2.3725852980716557e-07,
+      "loss": 0.7644,
+      "step": 2350
+    },
+    {
+      "epoch": 0.908653383386274,
+      "grad_norm": 0.22798039019107819,
+      "learning_rate": 2.1864844158503983e-07,
+      "loss": 0.7783,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9125036095870632,
+      "grad_norm": 0.2749248743057251,
+      "learning_rate": 2.007820373979702e-07,
+      "loss": 0.7712,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9163538357878526,
+      "grad_norm": 0.23876889050006866,
+      "learning_rate": 1.8366209615886532e-07,
+      "loss": 0.7778,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9202040619886418,
+      "grad_norm": 0.23598432540893555,
+      "learning_rate": 1.6729128067693613e-07,
+      "loss": 0.7909,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9240542881894311,
+      "grad_norm": 0.2578764259815216,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.7903,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9279045143902205,
+      "grad_norm": 0.23426567018032074,
+      "learning_rate": 1.3680709523610058e-07,
+      "loss": 0.7741,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9317547405910097,
+      "grad_norm": 0.22543291747570038,
+      "learning_rate": 1.2269846674030418e-07,
+      "loss": 0.794,
+      "step": 2420
+    },
+    {
+      "epoch": 0.935604966791799,
+      "grad_norm": 0.23952874541282654,
+      "learning_rate": 1.0934844619042595e-07,
+      "loss": 0.7927,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9394551929925883,
+      "grad_norm": 0.25108540058135986,
+      "learning_rate": 9.675911002803961e-08,
+      "loss": 0.7805,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9433054191933776,
+      "grad_norm": 0.2646595537662506,
+      "learning_rate": 8.493241637904381e-08,
+      "loss": 0.7767,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9471556453941669,
+      "grad_norm": 0.24272465705871582,
+      "learning_rate": 7.387020474909689e-08,
+      "loss": 0.7981,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9510058715949562,
+      "grad_norm": 0.26722830533981323,
+      "learning_rate": 6.357419573750856e-08,
+      "loss": 0.7912,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9548560977957455,
+      "grad_norm": 0.24296846985816956,
+      "learning_rate": 5.4045990769613365e-08,
+      "loss": 0.7756,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9587063239965348,
+      "grad_norm": 0.28137966990470886,
+      "learning_rate": 4.5287071847694434e-08,
+      "loss": 0.7792,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9625565501973241,
+      "grad_norm": 0.23351995646953583,
+      "learning_rate": 3.7298801320467817e-08,
+      "loss": 0.7747,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9664067763981133,
+      "grad_norm": 0.23447147011756897,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.7681,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9702570025989027,
+      "grad_norm": 0.2386709749698639,
+      "learning_rate": 2.3639055324410355e-08,
+      "loss": 0.7907,
+      "step": 2520
+    },
+    {
+      "epoch": 0.974107228799692,
+      "grad_norm": 0.24166277050971985,
+      "learning_rate": 1.7969704471369477e-08,
+      "loss": 0.7732,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9779574550004813,
+      "grad_norm": 0.22922277450561523,
+      "learning_rate": 1.3075250914142413e-08,
+      "loss": 0.7775,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9818076812012706,
+      "grad_norm": 0.23014265298843384,
+      "learning_rate": 8.956455928476049e-09,
+      "loss": 0.7867,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9856579074020598,
+      "grad_norm": 0.2459011822938919,
+      "learning_rate": 5.6139601453786275e-09,
+      "loss": 0.7773,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9895081336028492,
+      "grad_norm": 0.24084420502185822,
+      "learning_rate": 3.0482834514805602e-09,
+      "loss": 0.766,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9933583598036385,
+      "grad_norm": 0.22241507470607758,
+      "learning_rate": 1.2598249081713366e-09,
+      "loss": 0.7615,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9972085860044277,
+      "grad_norm": 0.2566074728965759,
+      "learning_rate": 2.488626895302826e-10,
+      "loss": 0.7925,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9999037443449803,
+      "step": 2597,
+      "total_flos": 3.652195211100029e+18,
+      "train_loss": 0.8174839934157003,
+      "train_runtime": 10312.6305,
+      "train_samples_per_second": 64.475,
+      "train_steps_per_second": 0.252
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2597,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.652195211100029e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3/mc2_m16_224/finetune/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25686806cb1da6cc140c9d394cf6bbf0fd93cd2de6d1441bf38f77547e678697
+size 6136

continuous/qwen3/mc2_m16_224/pretrain/config.json

@@ -0,0 +1,60 @@
+{
+  "_attn_implementation_autoset": true,
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "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,
+  "max_position_embeddings": 40960,
+  "max_window_layers": 28,
+  "mm_hidden_size": 512,
+  "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-M-16",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "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",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "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_m16_224.pt",
+  "vocab_size": 151936
+}

continuous/qwen3/mc2_m16_224/pretrain/mm_projector.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e27cd340b3c5180ed209261236fe1c4ab743898a4f7e7ea21c6700271fcedbbf
+size 20990008

continuous/qwen3/mc2_m16_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": 8.263691902160645,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 9.5278,
+      "step": 10
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 0.9499373435974121,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 5.6525,
+      "step": 20
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.1157476082444191,
+      "learning_rate": 0.0009999907688176173,
+      "loss": 4.7837,
+      "step": 30
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.0696597769856453,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 4.6809,
+      "step": 40
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.0800262838602066,
+      "learning_rate": 0.0009959345572940772,
+      "loss": 4.64,
+      "step": 50
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.05334581434726715,
+      "learning_rate": 0.0009911550081155983,
+      "loss": 4.6184,
+      "step": 60
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.045210205018520355,
+      "learning_rate": 0.0009845624347566433,
+      "loss": 4.587,
+      "step": 70
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.08525381237268448,
+      "learning_rate": 0.0009761811727013547,
+      "loss": 4.5676,
+      "step": 80
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.08886343240737915,
+      "learning_rate": 0.0009660421601067666,
+      "loss": 4.5322,
+      "step": 90
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.14597123861312866,
+      "learning_rate": 0.0009541828235992926,
+      "loss": 4.5087,
+      "step": 100
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.16878250241279602,
+      "learning_rate": 0.0009406469401200151,
+      "loss": 4.4802,
+      "step": 110
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.1627032309770584,
+      "learning_rate": 0.0009254844753287492,
+      "loss": 4.467,
+      "step": 120
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.12593714892864227,
+      "learning_rate": 0.0009087513991633924,
+      "loss": 4.4395,
+      "step": 130
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.17683015763759613,
+      "learning_rate": 0.0008905094792353917,
+      "loss": 4.4194,
+      "step": 140
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.1391962170600891,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 4.4178,
+      "step": 150
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.14064252376556396,
+      "learning_rate": 0.0008497737783108237,
+      "loss": 4.4027,
+      "step": 160
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.11863978207111359,
+      "learning_rate": 0.0008274303669726426,
+      "loss": 4.3896,
+      "step": 170
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.13673195242881775,
+      "learning_rate": 0.0008038782961218137,
+      "loss": 4.376,
+      "step": 180
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.13653913140296936,
+      "learning_rate": 0.000779204504653894,
+      "loss": 4.3717,
+      "step": 190
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.15684056282043457,
+      "learning_rate": 0.0007535000721258791,
+      "loss": 4.3525,
+      "step": 200
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.1207437664270401,
+      "learning_rate": 0.00072685988254984,
+      "loss": 4.3438,
+      "step": 210
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.15573497116565704,
+      "learning_rate": 0.0006993822741429906,
+      "loss": 4.3557,
+      "step": 220
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.1481485366821289,
+      "learning_rate": 0.0006711686763270818,
+      "loss": 4.3535,
+      "step": 230
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.1350519210100174,
+      "learning_rate": 0.0006423232353170831,
+      "loss": 4.3387,
+      "step": 240
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.13183097541332245,
+      "learning_rate": 0.0006129524296812335,
+      "loss": 4.318,
+      "step": 250
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.09445122629404068,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 4.3299,
+      "step": 260
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.12261876463890076,
+      "learning_rate": 0.0005530699351157809,
+      "loss": 4.3094,
+      "step": 270
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.12118884921073914,
+      "learning_rate": 0.0005227792933277943,
+      "loss": 4.3087,
+      "step": 280
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.11086063086986542,
+      "learning_rate": 0.0004924045652352092,
+      "loss": 4.2942,
+      "step": 290
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.15731115639209747,
+      "learning_rate": 0.0004620578745374633,
+      "loss": 4.2966,
+      "step": 300
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.14452946186065674,
+      "learning_rate": 0.0004318512414382059,
+      "loss": 4.3098,
+      "step": 310
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.11758684366941452,
+      "learning_rate": 0.0004018961691397148,
+      "loss": 4.2959,
+      "step": 320
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.11536116153001785,
+      "learning_rate": 0.0003723032322457458,
+      "loss": 4.2803,
+      "step": 330
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.12444861978292465,
+      "learning_rate": 0.0003431816685921625,
+      "loss": 4.2791,
+      "step": 340
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.11540855467319489,
+      "learning_rate": 0.0003146389760120416,
+      "loss": 4.2871,
+      "step": 350
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.12380729615688324,
+      "learning_rate": 0.00028678051552373485,
+      "loss": 4.2928,
+      "step": 360
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.10324236750602722,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 4.2763,
+      "step": 370
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.10322773456573486,
+      "learning_rate": 0.0002335247266004629,
+      "loss": 4.2921,
+      "step": 380
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.11596544831991196,
+      "learning_rate": 0.000208323983828828,
+      "loss": 4.271,
+      "step": 390
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.12099551409482956,
+      "learning_rate": 0.0001841999188095224,
+      "loss": 4.2747,
+      "step": 400
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.10912972688674927,
+      "learning_rate": 0.00016124158186777676,
+      "loss": 4.2766,
+      "step": 410
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.09746512770652771,
+      "learning_rate": 0.0001395337202205211,
+      "loss": 4.2672,
+      "step": 420
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.0992540642619133,
+      "learning_rate": 0.00011915646514489292,
+      "loss": 4.2729,
+      "step": 430
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.10315901041030884,
+      "learning_rate": 0.00010018503618578818,
+      "loss": 4.281,
+      "step": 440
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.10308006405830383,
+      "learning_rate": 8.268946349432583e-05,
+      "loss": 4.2697,
+      "step": 450
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.11523585766553879,
+      "learning_rate": 6.673432932217117e-05,
+      "loss": 4.2611,
+      "step": 460
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.10282143950462341,
+      "learning_rate": 5.2378529625954684e-05,
+      "loss": 4.2499,
+      "step": 470
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.09811653941869736,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 4.2613,
+      "step": 480
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.08894379436969757,
+      "learning_rate": 2.8670803372379538e-05,
+      "loss": 4.2629,
+      "step": 490
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.08747044205665588,
+      "learning_rate": 1.9406390288876586e-05,
+      "loss": 4.2735,
+      "step": 500
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.09227199107408524,
+      "learning_rate": 1.191601558631461e-05,
+      "loss": 4.2623,
+      "step": 510
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.08843649178743362,
+      "learning_rate": 6.2273288462559305e-06,
+      "loss": 4.2536,
+      "step": 520
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.09049606323242188,
+      "learning_rate": 2.361328992552314e-06,
+      "loss": 4.2581,
+      "step": 530
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.09209471195936203,
+      "learning_rate": 3.322867770044202e-07,
+      "loss": 4.2631,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "step": 545,
+      "total_flos": 2.889885108938998e+17,
+      "train_loss": 4.480545894377822,
+      "train_runtime": 5620.1513,
+      "train_samples_per_second": 99.308,
+      "train_steps_per_second": 0.097
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 545,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.889885108938998e+17,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3/mc2_m16_224_mt5/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/qwen3/mc2_m16_224_mt5/finetune/config.json

@@ -0,0 +1,59 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "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,
+  "max_position_embeddings": 40960,
+  "max_window_layers": 28,
+  "mm_hidden_size": 512,
+  "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-M-16",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "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",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "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_m16_224_mt5.pt",
+  "vocab_size": 151936
+}

continuous/qwen3/mc2_m16_224_mt5/finetune/generation_config.json

@@ -0,0 +1,13 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.6,
+  "top_k": 20,
+  "top_p": 0.95,
+  "transformers_version": "4.51.3"
+}