Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

config.json

@@ -0,0 +1,26 @@
+{
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 5120,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 1024000,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 40,
+  "num_key_value_heads": 8,
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "use_cache": false,
+  "vocab_size": 131072
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "transformers_version": "4.51.3"
+}

model-00001-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:719b52515e18e4187666f5e5cfe6297ca699308fbdb6fd8fdd795ef11bb0cb75
+size 4865522496

model-00002-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7470034a5cb1db5a69012715b2dc2953fe926875ef8e4a94d306c02af850c88
+size 4907529424

model-00003-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68fc611e2046025a65f31a8e8491965438aa0671ec66448375a8dbbc67999575
+size 4907529456

model-00004-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b807419c2174c381f1368277be7fa2f3337b3e83946cd5402dea841c70dcd2d
+size 4907529456

model-00005-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03527b3513f4aee24d1bb7969ba1748db8d1139e5c2c33c11f2d4a9a3d550ee2
+size 4907496272

model.safetensors.index.json

@@ -0,0 +1,370 @@
+{
+  "metadata": {
+    "total_size": 24495564800
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00005-of-00005.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.34.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.34.mlp.down_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.34.mlp.gate_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.34.mlp.up_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.34.self_attn.k_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.34.self_attn.o_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.34.self_attn.q_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.34.self_attn.v_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.35.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.35.mlp.down_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.35.mlp.gate_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.35.mlp.up_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.35.self_attn.k_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.35.self_attn.o_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.35.self_attn.q_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.35.self_attn.v_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.36.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.36.mlp.down_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.36.mlp.gate_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.36.mlp.up_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.36.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.36.self_attn.k_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.36.self_attn.o_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.36.self_attn.q_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.36.self_attn.v_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.37.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.37.mlp.down_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.37.mlp.gate_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.37.mlp.up_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.37.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.37.self_attn.k_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.37.self_attn.o_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.37.self_attn.q_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.37.self_attn.v_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.38.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.38.mlp.down_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.38.mlp.gate_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.38.mlp.up_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.38.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.38.self_attn.k_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.38.self_attn.o_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.38.self_attn.q_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.38.self_attn.v_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.39.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.39.mlp.down_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.39.mlp.gate_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.39.mlp.up_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.39.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.39.self_attn.k_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.39.self_attn.o_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.39.self_attn.q_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.39.self_attn.v_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.norm.weight": "model-00005-of-00005.safetensors"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0240ce510f08e6c2041724e9043e33be9d251d1e4a4d94eb68cd47b954b61d2
+size 17078292

trainer_state.json

@@ -0,0 +1,2134 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6954505940307157,
+  "eval_steps": 500,
+  "global_step": 300,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0023181686467690526,
+      "grad_norm": 28.501070022583008,
+      "learning_rate": 2e-05,
+      "loss": 2.6862,
+      "step": 1
+    },
+    {
+      "epoch": 0.004636337293538105,
+      "grad_norm": 2.4246978759765625,
+      "learning_rate": 1.998453209590101e-05,
+      "loss": 1.2503,
+      "step": 2
+    },
+    {
+      "epoch": 0.006954505940307157,
+      "grad_norm": 8.621129989624023,
+      "learning_rate": 1.996906419180201e-05,
+      "loss": 1.9701,
+      "step": 3
+    },
+    {
+      "epoch": 0.00927267458707621,
+      "grad_norm": 2.628929853439331,
+      "learning_rate": 1.995359628770302e-05,
+      "loss": 1.4985,
+      "step": 4
+    },
+    {
+      "epoch": 0.011590843233845263,
+      "grad_norm": 5.25876522064209,
+      "learning_rate": 1.993812838360402e-05,
+      "loss": 1.3525,
+      "step": 5
+    },
+    {
+      "epoch": 0.013909011880614315,
+      "grad_norm": 1.959221363067627,
+      "learning_rate": 1.992266047950503e-05,
+      "loss": 1.2634,
+      "step": 6
+    },
+    {
+      "epoch": 0.016227180527383367,
+      "grad_norm": 4.436636447906494,
+      "learning_rate": 1.9907192575406035e-05,
+      "loss": 1.3113,
+      "step": 7
+    },
+    {
+      "epoch": 0.01854534917415242,
+      "grad_norm": 2.8270750045776367,
+      "learning_rate": 1.989172467130704e-05,
+      "loss": 1.2802,
+      "step": 8
+    },
+    {
+      "epoch": 0.02086351782092147,
+      "grad_norm": 1.8319437503814697,
+      "learning_rate": 1.9876256767208045e-05,
+      "loss": 1.2364,
+      "step": 9
+    },
+    {
+      "epoch": 0.023181686467690525,
+      "grad_norm": 1.2635818719863892,
+      "learning_rate": 1.986078886310905e-05,
+      "loss": 1.2281,
+      "step": 10
+    },
+    {
+      "epoch": 0.025499855114459576,
+      "grad_norm": 1.3797353506088257,
+      "learning_rate": 1.9845320959010055e-05,
+      "loss": 1.2345,
+      "step": 11
+    },
+    {
+      "epoch": 0.02781802376122863,
+      "grad_norm": 1.020139455795288,
+      "learning_rate": 1.982985305491106e-05,
+      "loss": 1.2106,
+      "step": 12
+    },
+    {
+      "epoch": 0.030136192407997683,
+      "grad_norm": 1.2911367416381836,
+      "learning_rate": 1.9814385150812065e-05,
+      "loss": 1.2277,
+      "step": 13
+    },
+    {
+      "epoch": 0.032454361054766734,
+      "grad_norm": 1.878492832183838,
+      "learning_rate": 1.979891724671307e-05,
+      "loss": 1.1971,
+      "step": 14
+    },
+    {
+      "epoch": 0.034772529701535784,
+      "grad_norm": 1.455005407333374,
+      "learning_rate": 1.978344934261408e-05,
+      "loss": 1.1909,
+      "step": 15
+    },
+    {
+      "epoch": 0.03709069834830484,
+      "grad_norm": 1.5538443326950073,
+      "learning_rate": 1.976798143851508e-05,
+      "loss": 1.227,
+      "step": 16
+    },
+    {
+      "epoch": 0.03940886699507389,
+      "grad_norm": 1.4034634828567505,
+      "learning_rate": 1.975251353441609e-05,
+      "loss": 1.1373,
+      "step": 17
+    },
+    {
+      "epoch": 0.04172703564184294,
+      "grad_norm": 1.3893576860427856,
+      "learning_rate": 1.973704563031709e-05,
+      "loss": 1.178,
+      "step": 18
+    },
+    {
+      "epoch": 0.044045204288612,
+      "grad_norm": 1.352433204650879,
+      "learning_rate": 1.97215777262181e-05,
+      "loss": 1.163,
+      "step": 19
+    },
+    {
+      "epoch": 0.04636337293538105,
+      "grad_norm": 1.179675579071045,
+      "learning_rate": 1.9706109822119105e-05,
+      "loss": 1.1223,
+      "step": 20
+    },
+    {
+      "epoch": 0.0486815415821501,
+      "grad_norm": 1.208991289138794,
+      "learning_rate": 1.9690641918020112e-05,
+      "loss": 1.1867,
+      "step": 21
+    },
+    {
+      "epoch": 0.05099971022891915,
+      "grad_norm": 0.8430147171020508,
+      "learning_rate": 1.9675174013921115e-05,
+      "loss": 1.1488,
+      "step": 22
+    },
+    {
+      "epoch": 0.05331787887568821,
+      "grad_norm": 0.7868801355361938,
+      "learning_rate": 1.9659706109822122e-05,
+      "loss": 1.1035,
+      "step": 23
+    },
+    {
+      "epoch": 0.05563604752245726,
+      "grad_norm": 0.9440180063247681,
+      "learning_rate": 1.9644238205723125e-05,
+      "loss": 1.1582,
+      "step": 24
+    },
+    {
+      "epoch": 0.05795421616922631,
+      "grad_norm": 1.0409374237060547,
+      "learning_rate": 1.9628770301624132e-05,
+      "loss": 1.1304,
+      "step": 25
+    },
+    {
+      "epoch": 0.06027238481599537,
+      "grad_norm": 1.4376304149627686,
+      "learning_rate": 1.961330239752514e-05,
+      "loss": 1.1535,
+      "step": 26
+    },
+    {
+      "epoch": 0.06259055346276442,
+      "grad_norm": 0.7775300741195679,
+      "learning_rate": 1.9597834493426142e-05,
+      "loss": 1.0941,
+      "step": 27
+    },
+    {
+      "epoch": 0.06490872210953347,
+      "grad_norm": 1.294155478477478,
+      "learning_rate": 1.958236658932715e-05,
+      "loss": 1.1859,
+      "step": 28
+    },
+    {
+      "epoch": 0.06722689075630252,
+      "grad_norm": 1.0435543060302734,
+      "learning_rate": 1.9566898685228152e-05,
+      "loss": 1.147,
+      "step": 29
+    },
+    {
+      "epoch": 0.06954505940307157,
+      "grad_norm": 1.1211307048797607,
+      "learning_rate": 1.955143078112916e-05,
+      "loss": 1.1148,
+      "step": 30
+    },
+    {
+      "epoch": 0.07186322804984063,
+      "grad_norm": 0.9777933955192566,
+      "learning_rate": 1.9535962877030165e-05,
+      "loss": 1.1266,
+      "step": 31
+    },
+    {
+      "epoch": 0.07418139669660968,
+      "grad_norm": 0.9920445680618286,
+      "learning_rate": 1.952049497293117e-05,
+      "loss": 1.1535,
+      "step": 32
+    },
+    {
+      "epoch": 0.07649956534337873,
+      "grad_norm": 0.7202315330505371,
+      "learning_rate": 1.9505027068832175e-05,
+      "loss": 1.139,
+      "step": 33
+    },
+    {
+      "epoch": 0.07881773399014778,
+      "grad_norm": 0.6207343339920044,
+      "learning_rate": 1.9489559164733182e-05,
+      "loss": 1.1105,
+      "step": 34
+    },
+    {
+      "epoch": 0.08113590263691683,
+      "grad_norm": 1.078873634338379,
+      "learning_rate": 1.9474091260634185e-05,
+      "loss": 1.1407,
+      "step": 35
+    },
+    {
+      "epoch": 0.08345407128368589,
+      "grad_norm": 0.9423937201499939,
+      "learning_rate": 1.9458623356535192e-05,
+      "loss": 1.145,
+      "step": 36
+    },
+    {
+      "epoch": 0.08577223993045494,
+      "grad_norm": 0.8272204399108887,
+      "learning_rate": 1.9443155452436195e-05,
+      "loss": 1.1141,
+      "step": 37
+    },
+    {
+      "epoch": 0.088090408577224,
+      "grad_norm": 0.9147409796714783,
+      "learning_rate": 1.9427687548337202e-05,
+      "loss": 1.1113,
+      "step": 38
+    },
+    {
+      "epoch": 0.09040857722399305,
+      "grad_norm": 1.4252681732177734,
+      "learning_rate": 1.941221964423821e-05,
+      "loss": 1.1633,
+      "step": 39
+    },
+    {
+      "epoch": 0.0927267458707621,
+      "grad_norm": 0.8701033592224121,
+      "learning_rate": 1.9396751740139212e-05,
+      "loss": 1.1243,
+      "step": 40
+    },
+    {
+      "epoch": 0.09504491451753115,
+      "grad_norm": 0.9681833386421204,
+      "learning_rate": 1.938128383604022e-05,
+      "loss": 1.1518,
+      "step": 41
+    },
+    {
+      "epoch": 0.0973630831643002,
+      "grad_norm": 1.1157395839691162,
+      "learning_rate": 1.9365815931941222e-05,
+      "loss": 1.1178,
+      "step": 42
+    },
+    {
+      "epoch": 0.09968125181106925,
+      "grad_norm": 0.7797672152519226,
+      "learning_rate": 1.935034802784223e-05,
+      "loss": 1.0887,
+      "step": 43
+    },
+    {
+      "epoch": 0.1019994204578383,
+      "grad_norm": 1.3890780210494995,
+      "learning_rate": 1.9334880123743235e-05,
+      "loss": 1.1307,
+      "step": 44
+    },
+    {
+      "epoch": 0.10431758910460737,
+      "grad_norm": 0.8613622784614563,
+      "learning_rate": 1.9319412219644242e-05,
+      "loss": 1.1662,
+      "step": 45
+    },
+    {
+      "epoch": 0.10663575775137642,
+      "grad_norm": 1.0830810070037842,
+      "learning_rate": 1.9303944315545245e-05,
+      "loss": 1.1796,
+      "step": 46
+    },
+    {
+      "epoch": 0.10895392639814547,
+      "grad_norm": 0.8859057426452637,
+      "learning_rate": 1.9288476411446252e-05,
+      "loss": 1.1584,
+      "step": 47
+    },
+    {
+      "epoch": 0.11127209504491452,
+      "grad_norm": 0.6870171427726746,
+      "learning_rate": 1.9273008507347255e-05,
+      "loss": 1.1035,
+      "step": 48
+    },
+    {
+      "epoch": 0.11359026369168357,
+      "grad_norm": 1.181731104850769,
+      "learning_rate": 1.9257540603248262e-05,
+      "loss": 1.1241,
+      "step": 49
+    },
+    {
+      "epoch": 0.11590843233845262,
+      "grad_norm": 1.8238871097564697,
+      "learning_rate": 1.924207269914927e-05,
+      "loss": 1.2359,
+      "step": 50
+    },
+    {
+      "epoch": 0.11822660098522167,
+      "grad_norm": 0.9966158270835876,
+      "learning_rate": 1.9226604795050272e-05,
+      "loss": 1.1387,
+      "step": 51
+    },
+    {
+      "epoch": 0.12054476963199073,
+      "grad_norm": 0.9722117185592651,
+      "learning_rate": 1.921113689095128e-05,
+      "loss": 1.1725,
+      "step": 52
+    },
+    {
+      "epoch": 0.12286293827875978,
+      "grad_norm": 1.1024895906448364,
+      "learning_rate": 1.9195668986852282e-05,
+      "loss": 1.1424,
+      "step": 53
+    },
+    {
+      "epoch": 0.12518110692552883,
+      "grad_norm": 1.157772421836853,
+      "learning_rate": 1.918020108275329e-05,
+      "loss": 1.1627,
+      "step": 54
+    },
+    {
+      "epoch": 0.12749927557229787,
+      "grad_norm": 1.0060945749282837,
+      "learning_rate": 1.9164733178654292e-05,
+      "loss": 1.1714,
+      "step": 55
+    },
+    {
+      "epoch": 0.12981744421906694,
+      "grad_norm": 1.3089208602905273,
+      "learning_rate": 1.91492652745553e-05,
+      "loss": 1.1394,
+      "step": 56
+    },
+    {
+      "epoch": 0.132135612865836,
+      "grad_norm": 1.0314549207687378,
+      "learning_rate": 1.9133797370456305e-05,
+      "loss": 1.1214,
+      "step": 57
+    },
+    {
+      "epoch": 0.13445378151260504,
+      "grad_norm": 0.848561704158783,
+      "learning_rate": 1.9118329466357312e-05,
+      "loss": 1.1364,
+      "step": 58
+    },
+    {
+      "epoch": 0.1367719501593741,
+      "grad_norm": 1.1316096782684326,
+      "learning_rate": 1.9102861562258315e-05,
+      "loss": 1.0673,
+      "step": 59
+    },
+    {
+      "epoch": 0.13909011880614314,
+      "grad_norm": 0.910849928855896,
+      "learning_rate": 1.9087393658159322e-05,
+      "loss": 1.0848,
+      "step": 60
+    },
+    {
+      "epoch": 0.1414082874529122,
+      "grad_norm": 1.2117191553115845,
+      "learning_rate": 1.9071925754060325e-05,
+      "loss": 1.2018,
+      "step": 61
+    },
+    {
+      "epoch": 0.14372645609968127,
+      "grad_norm": 1.3396297693252563,
+      "learning_rate": 1.9056457849961332e-05,
+      "loss": 1.1773,
+      "step": 62
+    },
+    {
+      "epoch": 0.1460446247464503,
+      "grad_norm": 1.0479182004928589,
+      "learning_rate": 1.904098994586234e-05,
+      "loss": 1.1557,
+      "step": 63
+    },
+    {
+      "epoch": 0.14836279339321937,
+      "grad_norm": 1.0438235998153687,
+      "learning_rate": 1.9025522041763342e-05,
+      "loss": 1.0985,
+      "step": 64
+    },
+    {
+      "epoch": 0.1506809620399884,
+      "grad_norm": 1.3567838668823242,
+      "learning_rate": 1.901005413766435e-05,
+      "loss": 1.0874,
+      "step": 65
+    },
+    {
+      "epoch": 0.15299913068675747,
+      "grad_norm": 1.0052990913391113,
+      "learning_rate": 1.8994586233565352e-05,
+      "loss": 1.1008,
+      "step": 66
+    },
+    {
+      "epoch": 0.1553172993335265,
+      "grad_norm": 1.06718111038208,
+      "learning_rate": 1.897911832946636e-05,
+      "loss": 1.1004,
+      "step": 67
+    },
+    {
+      "epoch": 0.15763546798029557,
+      "grad_norm": 1.326567530632019,
+      "learning_rate": 1.8963650425367365e-05,
+      "loss": 1.1033,
+      "step": 68
+    },
+    {
+      "epoch": 0.15995363662706463,
+      "grad_norm": 1.1070104837417603,
+      "learning_rate": 1.894818252126837e-05,
+      "loss": 1.1287,
+      "step": 69
+    },
+    {
+      "epoch": 0.16227180527383367,
+      "grad_norm": 1.0842565298080444,
+      "learning_rate": 1.8932714617169375e-05,
+      "loss": 1.1005,
+      "step": 70
+    },
+    {
+      "epoch": 0.16458997392060273,
+      "grad_norm": 0.9498984813690186,
+      "learning_rate": 1.8917246713070382e-05,
+      "loss": 1.0714,
+      "step": 71
+    },
+    {
+      "epoch": 0.16690814256737177,
+      "grad_norm": 1.0374786853790283,
+      "learning_rate": 1.8901778808971385e-05,
+      "loss": 1.1789,
+      "step": 72
+    },
+    {
+      "epoch": 0.16922631121414083,
+      "grad_norm": 0.8770291209220886,
+      "learning_rate": 1.8886310904872392e-05,
+      "loss": 1.087,
+      "step": 73
+    },
+    {
+      "epoch": 0.17154447986090987,
+      "grad_norm": 1.4603461027145386,
+      "learning_rate": 1.88708430007734e-05,
+      "loss": 1.1119,
+      "step": 74
+    },
+    {
+      "epoch": 0.17386264850767894,
+      "grad_norm": 1.0779688358306885,
+      "learning_rate": 1.8855375096674402e-05,
+      "loss": 1.1346,
+      "step": 75
+    },
+    {
+      "epoch": 0.176180817154448,
+      "grad_norm": 1.1575367450714111,
+      "learning_rate": 1.883990719257541e-05,
+      "loss": 1.043,
+      "step": 76
+    },
+    {
+      "epoch": 0.17849898580121704,
+      "grad_norm": 0.991324245929718,
+      "learning_rate": 1.8824439288476412e-05,
+      "loss": 1.0807,
+      "step": 77
+    },
+    {
+      "epoch": 0.1808171544479861,
+      "grad_norm": 1.2354373931884766,
+      "learning_rate": 1.880897138437742e-05,
+      "loss": 1.0485,
+      "step": 78
+    },
+    {
+      "epoch": 0.18313532309475514,
+      "grad_norm": 1.3966253995895386,
+      "learning_rate": 1.8793503480278422e-05,
+      "loss": 1.0349,
+      "step": 79
+    },
+    {
+      "epoch": 0.1854534917415242,
+      "grad_norm": 1.1162493228912354,
+      "learning_rate": 1.877803557617943e-05,
+      "loss": 1.1038,
+      "step": 80
+    },
+    {
+      "epoch": 0.18777166038829324,
+      "grad_norm": 1.117507815361023,
+      "learning_rate": 1.8762567672080435e-05,
+      "loss": 1.0844,
+      "step": 81
+    },
+    {
+      "epoch": 0.1900898290350623,
+      "grad_norm": 1.083954930305481,
+      "learning_rate": 1.8747099767981442e-05,
+      "loss": 1.1057,
+      "step": 82
+    },
+    {
+      "epoch": 0.19240799768183137,
+      "grad_norm": 1.0554256439208984,
+      "learning_rate": 1.8731631863882445e-05,
+      "loss": 1.0824,
+      "step": 83
+    },
+    {
+      "epoch": 0.1947261663286004,
+      "grad_norm": 1.1306427717208862,
+      "learning_rate": 1.8716163959783452e-05,
+      "loss": 1.0552,
+      "step": 84
+    },
+    {
+      "epoch": 0.19704433497536947,
+      "grad_norm": 1.1089762449264526,
+      "learning_rate": 1.8700696055684455e-05,
+      "loss": 1.0374,
+      "step": 85
+    },
+    {
+      "epoch": 0.1993625036221385,
+      "grad_norm": 1.0922352075576782,
+      "learning_rate": 1.8685228151585462e-05,
+      "loss": 1.2775,
+      "step": 86
+    },
+    {
+      "epoch": 0.20168067226890757,
+      "grad_norm": 4.459912300109863,
+      "learning_rate": 1.866976024748647e-05,
+      "loss": 1.061,
+      "step": 87
+    },
+    {
+      "epoch": 0.2039988409156766,
+      "grad_norm": 0.9974443316459656,
+      "learning_rate": 1.8654292343387472e-05,
+      "loss": 1.0759,
+      "step": 88
+    },
+    {
+      "epoch": 0.20631700956244567,
+      "grad_norm": 0.914336621761322,
+      "learning_rate": 1.863882443928848e-05,
+      "loss": 1.0549,
+      "step": 89
+    },
+    {
+      "epoch": 0.20863517820921473,
+      "grad_norm": 0.8654055595397949,
+      "learning_rate": 1.8623356535189482e-05,
+      "loss": 1.0634,
+      "step": 90
+    },
+    {
+      "epoch": 0.21095334685598377,
+      "grad_norm": 0.7419248819351196,
+      "learning_rate": 1.860788863109049e-05,
+      "loss": 1.0512,
+      "step": 91
+    },
+    {
+      "epoch": 0.21327151550275283,
+      "grad_norm": 1.4844622611999512,
+      "learning_rate": 1.8592420726991492e-05,
+      "loss": 1.0969,
+      "step": 92
+    },
+    {
+      "epoch": 0.21558968414952187,
+      "grad_norm": 1.29688560962677,
+      "learning_rate": 1.85769528228925e-05,
+      "loss": 1.0711,
+      "step": 93
+    },
+    {
+      "epoch": 0.21790785279629094,
+      "grad_norm": 1.421087622642517,
+      "learning_rate": 1.8561484918793505e-05,
+      "loss": 1.0772,
+      "step": 94
+    },
+    {
+      "epoch": 0.22022602144305997,
+      "grad_norm": 1.200110673904419,
+      "learning_rate": 1.8546017014694512e-05,
+      "loss": 1.1029,
+      "step": 95
+    },
+    {
+      "epoch": 0.22254419008982904,
+      "grad_norm": 1.025266408920288,
+      "learning_rate": 1.8530549110595515e-05,
+      "loss": 1.1009,
+      "step": 96
+    },
+    {
+      "epoch": 0.2248623587365981,
+      "grad_norm": 1.1353425979614258,
+      "learning_rate": 1.8515081206496522e-05,
+      "loss": 1.1109,
+      "step": 97
+    },
+    {
+      "epoch": 0.22718052738336714,
+      "grad_norm": 1.0217199325561523,
+      "learning_rate": 1.849961330239753e-05,
+      "loss": 1.1158,
+      "step": 98
+    },
+    {
+      "epoch": 0.2294986960301362,
+      "grad_norm": 0.9707551598548889,
+      "learning_rate": 1.8484145398298532e-05,
+      "loss": 1.0189,
+      "step": 99
+    },
+    {
+      "epoch": 0.23181686467690524,
+      "grad_norm": 0.8363978266716003,
+      "learning_rate": 1.846867749419954e-05,
+      "loss": 1.086,
+      "step": 100
+    },
+    {
+      "epoch": 0.2341350333236743,
+      "grad_norm": 1.1158548593521118,
+      "learning_rate": 1.8453209590100542e-05,
+      "loss": 1.0169,
+      "step": 101
+    },
+    {
+      "epoch": 0.23645320197044334,
+      "grad_norm": 0.8453333973884583,
+      "learning_rate": 1.843774168600155e-05,
+      "loss": 1.0584,
+      "step": 102
+    },
+    {
+      "epoch": 0.2387713706172124,
+      "grad_norm": 1.572540044784546,
+      "learning_rate": 1.8422273781902552e-05,
+      "loss": 1.105,
+      "step": 103
+    },
+    {
+      "epoch": 0.24108953926398147,
+      "grad_norm": 0.9463809132575989,
+      "learning_rate": 1.840680587780356e-05,
+      "loss": 1.0552,
+      "step": 104
+    },
+    {
+      "epoch": 0.2434077079107505,
+      "grad_norm": 0.8801397681236267,
+      "learning_rate": 1.8391337973704565e-05,
+      "loss": 1.0615,
+      "step": 105
+    },
+    {
+      "epoch": 0.24572587655751957,
+      "grad_norm": 1.251951813697815,
+      "learning_rate": 1.837587006960557e-05,
+      "loss": 1.1217,
+      "step": 106
+    },
+    {
+      "epoch": 0.2480440452042886,
+      "grad_norm": 1.297305703163147,
+      "learning_rate": 1.8360402165506575e-05,
+      "loss": 1.1102,
+      "step": 107
+    },
+    {
+      "epoch": 0.25036221385105767,
+      "grad_norm": 0.9023735523223877,
+      "learning_rate": 1.8344934261407582e-05,
+      "loss": 1.0786,
+      "step": 108
+    },
+    {
+      "epoch": 0.2526803824978267,
+      "grad_norm": 0.7729614973068237,
+      "learning_rate": 1.8329466357308585e-05,
+      "loss": 1.0636,
+      "step": 109
+    },
+    {
+      "epoch": 0.25499855114459574,
+      "grad_norm": 1.458217978477478,
+      "learning_rate": 1.8313998453209592e-05,
+      "loss": 1.0998,
+      "step": 110
+    },
+    {
+      "epoch": 0.25731671979136483,
+      "grad_norm": 1.295067548751831,
+      "learning_rate": 1.82985305491106e-05,
+      "loss": 1.0756,
+      "step": 111
+    },
+    {
+      "epoch": 0.25963488843813387,
+      "grad_norm": 0.7502389550209045,
+      "learning_rate": 1.8283062645011602e-05,
+      "loss": 1.048,
+      "step": 112
+    },
+    {
+      "epoch": 0.2619530570849029,
+      "grad_norm": 0.7939056158065796,
+      "learning_rate": 1.826759474091261e-05,
+      "loss": 1.1759,
+      "step": 113
+    },
+    {
+      "epoch": 0.264271225731672,
+      "grad_norm": 0.8996245861053467,
+      "learning_rate": 1.8252126836813612e-05,
+      "loss": 1.0799,
+      "step": 114
+    },
+    {
+      "epoch": 0.26658939437844104,
+      "grad_norm": 1.1998958587646484,
+      "learning_rate": 1.823665893271462e-05,
+      "loss": 1.0622,
+      "step": 115
+    },
+    {
+      "epoch": 0.2689075630252101,
+      "grad_norm": 0.7442781329154968,
+      "learning_rate": 1.8221191028615622e-05,
+      "loss": 1.1025,
+      "step": 116
+    },
+    {
+      "epoch": 0.2712257316719791,
+      "grad_norm": 1.0310958623886108,
+      "learning_rate": 1.820572312451663e-05,
+      "loss": 1.0557,
+      "step": 117
+    },
+    {
+      "epoch": 0.2735439003187482,
+      "grad_norm": 1.071614384651184,
+      "learning_rate": 1.8190255220417635e-05,
+      "loss": 1.1166,
+      "step": 118
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 0.8248165845870972,
+      "learning_rate": 1.8174787316318642e-05,
+      "loss": 1.1389,
+      "step": 119
+    },
+    {
+      "epoch": 0.2781802376122863,
+      "grad_norm": 0.9798437356948853,
+      "learning_rate": 1.8159319412219645e-05,
+      "loss": 1.0331,
+      "step": 120
+    },
+    {
+      "epoch": 0.28049840625905537,
+      "grad_norm": 0.7473148703575134,
+      "learning_rate": 1.8143851508120652e-05,
+      "loss": 1.0211,
+      "step": 121
+    },
+    {
+      "epoch": 0.2828165749058244,
+      "grad_norm": 1.3669893741607666,
+      "learning_rate": 1.812838360402166e-05,
+      "loss": 1.1495,
+      "step": 122
+    },
+    {
+      "epoch": 0.28513474355259344,
+      "grad_norm": 0.7082749009132385,
+      "learning_rate": 1.8112915699922662e-05,
+      "loss": 1.0244,
+      "step": 123
+    },
+    {
+      "epoch": 0.28745291219936253,
+      "grad_norm": 1.2198262214660645,
+      "learning_rate": 1.809744779582367e-05,
+      "loss": 1.1052,
+      "step": 124
+    },
+    {
+      "epoch": 0.28977108084613157,
+      "grad_norm": 2.2517826557159424,
+      "learning_rate": 1.8081979891724672e-05,
+      "loss": 1.0324,
+      "step": 125
+    },
+    {
+      "epoch": 0.2920892494929006,
+      "grad_norm": 0.9831060767173767,
+      "learning_rate": 1.806651198762568e-05,
+      "loss": 1.1176,
+      "step": 126
+    },
+    {
+      "epoch": 0.29440741813966964,
+      "grad_norm": 0.8122763633728027,
+      "learning_rate": 1.8051044083526682e-05,
+      "loss": 1.0238,
+      "step": 127
+    },
+    {
+      "epoch": 0.29672558678643873,
+      "grad_norm": 1.0002597570419312,
+      "learning_rate": 1.803557617942769e-05,
+      "loss": 1.0847,
+      "step": 128
+    },
+    {
+      "epoch": 0.29904375543320777,
+      "grad_norm": 0.8262125253677368,
+      "learning_rate": 1.8020108275328692e-05,
+      "loss": 1.1149,
+      "step": 129
+    },
+    {
+      "epoch": 0.3013619240799768,
+      "grad_norm": 1.2185602188110352,
+      "learning_rate": 1.80046403712297e-05,
+      "loss": 1.0688,
+      "step": 130
+    },
+    {
+      "epoch": 0.3036800927267459,
+      "grad_norm": 1.2163466215133667,
+      "learning_rate": 1.7989172467130705e-05,
+      "loss": 1.0366,
+      "step": 131
+    },
+    {
+      "epoch": 0.30599826137351493,
+      "grad_norm": 0.6977062225341797,
+      "learning_rate": 1.7973704563031712e-05,
+      "loss": 1.0853,
+      "step": 132
+    },
+    {
+      "epoch": 0.30831643002028397,
+      "grad_norm": 0.8096152544021606,
+      "learning_rate": 1.7958236658932715e-05,
+      "loss": 1.0104,
+      "step": 133
+    },
+    {
+      "epoch": 0.310634598667053,
+      "grad_norm": 0.9192125201225281,
+      "learning_rate": 1.7942768754833722e-05,
+      "loss": 1.0596,
+      "step": 134
+    },
+    {
+      "epoch": 0.3129527673138221,
+      "grad_norm": 0.7752702236175537,
+      "learning_rate": 1.792730085073473e-05,
+      "loss": 1.1281,
+      "step": 135
+    },
+    {
+      "epoch": 0.31527093596059114,
+      "grad_norm": 1.0385123491287231,
+      "learning_rate": 1.7911832946635732e-05,
+      "loss": 1.0857,
+      "step": 136
+    },
+    {
+      "epoch": 0.3175891046073602,
+      "grad_norm": 2.8603484630584717,
+      "learning_rate": 1.789636504253674e-05,
+      "loss": 1.1425,
+      "step": 137
+    },
+    {
+      "epoch": 0.31990727325412927,
+      "grad_norm": 1.104943037033081,
+      "learning_rate": 1.7880897138437742e-05,
+      "loss": 1.1464,
+      "step": 138
+    },
+    {
+      "epoch": 0.3222254419008983,
+      "grad_norm": 1.7259231805801392,
+      "learning_rate": 1.786542923433875e-05,
+      "loss": 1.0472,
+      "step": 139
+    },
+    {
+      "epoch": 0.32454361054766734,
+      "grad_norm": 0.9174676537513733,
+      "learning_rate": 1.7849961330239752e-05,
+      "loss": 1.0523,
+      "step": 140
+    },
+    {
+      "epoch": 0.3268617791944364,
+      "grad_norm": 0.9572336673736572,
+      "learning_rate": 1.783449342614076e-05,
+      "loss": 1.0327,
+      "step": 141
+    },
+    {
+      "epoch": 0.32917994784120547,
+      "grad_norm": 0.6567716598510742,
+      "learning_rate": 1.7819025522041766e-05,
+      "loss": 1.0979,
+      "step": 142
+    },
+    {
+      "epoch": 0.3314981164879745,
+      "grad_norm": 1.8695584535598755,
+      "learning_rate": 1.780355761794277e-05,
+      "loss": 1.0659,
+      "step": 143
+    },
+    {
+      "epoch": 0.33381628513474354,
+      "grad_norm": 0.8160743713378906,
+      "learning_rate": 1.7788089713843776e-05,
+      "loss": 1.0746,
+      "step": 144
+    },
+    {
+      "epoch": 0.33613445378151263,
+      "grad_norm": 0.7144508957862854,
+      "learning_rate": 1.7772621809744782e-05,
+      "loss": 1.002,
+      "step": 145
+    },
+    {
+      "epoch": 0.33845262242828167,
+      "grad_norm": 0.8914051055908203,
+      "learning_rate": 1.7757153905645786e-05,
+      "loss": 1.0334,
+      "step": 146
+    },
+    {
+      "epoch": 0.3407707910750507,
+      "grad_norm": 1.1182371377944946,
+      "learning_rate": 1.7741686001546792e-05,
+      "loss": 1.0535,
+      "step": 147
+    },
+    {
+      "epoch": 0.34308895972181974,
+      "grad_norm": 0.6911827325820923,
+      "learning_rate": 1.77262180974478e-05,
+      "loss": 1.0783,
+      "step": 148
+    },
+    {
+      "epoch": 0.34540712836858883,
+      "grad_norm": 1.141491413116455,
+      "learning_rate": 1.7710750193348802e-05,
+      "loss": 1.0522,
+      "step": 149
+    },
+    {
+      "epoch": 0.34772529701535787,
+      "grad_norm": 1.103798747062683,
+      "learning_rate": 1.769528228924981e-05,
+      "loss": 1.1178,
+      "step": 150
+    },
+    {
+      "epoch": 0.3500434656621269,
+      "grad_norm": 1.1297893524169922,
+      "learning_rate": 1.7679814385150812e-05,
+      "loss": 1.19,
+      "step": 151
+    },
+    {
+      "epoch": 0.352361634308896,
+      "grad_norm": 0.8850527405738831,
+      "learning_rate": 1.766434648105182e-05,
+      "loss": 0.9929,
+      "step": 152
+    },
+    {
+      "epoch": 0.35467980295566504,
+      "grad_norm": 1.096604585647583,
+      "learning_rate": 1.7648878576952822e-05,
+      "loss": 1.0992,
+      "step": 153
+    },
+    {
+      "epoch": 0.35699797160243407,
+      "grad_norm": 0.9644438624382019,
+      "learning_rate": 1.763341067285383e-05,
+      "loss": 1.0421,
+      "step": 154
+    },
+    {
+      "epoch": 0.3593161402492031,
+      "grad_norm": 1.0480139255523682,
+      "learning_rate": 1.7617942768754836e-05,
+      "loss": 1.0495,
+      "step": 155
+    },
+    {
+      "epoch": 0.3616343088959722,
+      "grad_norm": 3.11247181892395,
+      "learning_rate": 1.7602474864655842e-05,
+      "loss": 1.0407,
+      "step": 156
+    },
+    {
+      "epoch": 0.36395247754274124,
+      "grad_norm": 0.9178793430328369,
+      "learning_rate": 1.7587006960556846e-05,
+      "loss": 1.0208,
+      "step": 157
+    },
+    {
+      "epoch": 0.3662706461895103,
+      "grad_norm": 1.0008949041366577,
+      "learning_rate": 1.7571539056457852e-05,
+      "loss": 1.036,
+      "step": 158
+    },
+    {
+      "epoch": 0.36858881483627937,
+      "grad_norm": 0.9334746599197388,
+      "learning_rate": 1.755607115235886e-05,
+      "loss": 1.1097,
+      "step": 159
+    },
+    {
+      "epoch": 0.3709069834830484,
+      "grad_norm": 0.9296855330467224,
+      "learning_rate": 1.7540603248259862e-05,
+      "loss": 1.0597,
+      "step": 160
+    },
+    {
+      "epoch": 0.37322515212981744,
+      "grad_norm": 0.7528197765350342,
+      "learning_rate": 1.752513534416087e-05,
+      "loss": 1.0252,
+      "step": 161
+    },
+    {
+      "epoch": 0.3755433207765865,
+      "grad_norm": 0.7995203733444214,
+      "learning_rate": 1.7509667440061872e-05,
+      "loss": 1.0446,
+      "step": 162
+    },
+    {
+      "epoch": 0.37786148942335557,
+      "grad_norm": 0.7773709297180176,
+      "learning_rate": 1.749419953596288e-05,
+      "loss": 1.0527,
+      "step": 163
+    },
+    {
+      "epoch": 0.3801796580701246,
+      "grad_norm": 0.9108691811561584,
+      "learning_rate": 1.7478731631863882e-05,
+      "loss": 1.0571,
+      "step": 164
+    },
+    {
+      "epoch": 0.38249782671689364,
+      "grad_norm": 0.8565751910209656,
+      "learning_rate": 1.746326372776489e-05,
+      "loss": 1.1003,
+      "step": 165
+    },
+    {
+      "epoch": 0.38481599536366273,
+      "grad_norm": 1.3683419227600098,
+      "learning_rate": 1.7447795823665896e-05,
+      "loss": 1.0697,
+      "step": 166
+    },
+    {
+      "epoch": 0.38713416401043177,
+      "grad_norm": 0.6560070514678955,
+      "learning_rate": 1.74323279195669e-05,
+      "loss": 0.9556,
+      "step": 167
+    },
+    {
+      "epoch": 0.3894523326572008,
+      "grad_norm": 1.370263934135437,
+      "learning_rate": 1.7416860015467906e-05,
+      "loss": 1.0848,
+      "step": 168
+    },
+    {
+      "epoch": 0.39177050130396984,
+      "grad_norm": 1.013763427734375,
+      "learning_rate": 1.7401392111368912e-05,
+      "loss": 1.0643,
+      "step": 169
+    },
+    {
+      "epoch": 0.39408866995073893,
+      "grad_norm": 0.8194316029548645,
+      "learning_rate": 1.7385924207269916e-05,
+      "loss": 1.0708,
+      "step": 170
+    },
+    {
+      "epoch": 0.39640683859750797,
+      "grad_norm": 0.9241949319839478,
+      "learning_rate": 1.7370456303170922e-05,
+      "loss": 1.0502,
+      "step": 171
+    },
+    {
+      "epoch": 0.398725007244277,
+      "grad_norm": 0.9724448323249817,
+      "learning_rate": 1.735498839907193e-05,
+      "loss": 1.0152,
+      "step": 172
+    },
+    {
+      "epoch": 0.4010431758910461,
+      "grad_norm": 0.6559419631958008,
+      "learning_rate": 1.7339520494972932e-05,
+      "loss": 1.0148,
+      "step": 173
+    },
+    {
+      "epoch": 0.40336134453781514,
+      "grad_norm": 1.1617038249969482,
+      "learning_rate": 1.732405259087394e-05,
+      "loss": 1.0972,
+      "step": 174
+    },
+    {
+      "epoch": 0.4056795131845842,
+      "grad_norm": 0.7249406576156616,
+      "learning_rate": 1.7308584686774942e-05,
+      "loss": 0.9664,
+      "step": 175
+    },
+    {
+      "epoch": 0.4079976818313532,
+      "grad_norm": 1.2099742889404297,
+      "learning_rate": 1.729311678267595e-05,
+      "loss": 1.0314,
+      "step": 176
+    },
+    {
+      "epoch": 0.4103158504781223,
+      "grad_norm": 0.8690075278282166,
+      "learning_rate": 1.7277648878576952e-05,
+      "loss": 1.0789,
+      "step": 177
+    },
+    {
+      "epoch": 0.41263401912489134,
+      "grad_norm": 0.7662826180458069,
+      "learning_rate": 1.726218097447796e-05,
+      "loss": 1.014,
+      "step": 178
+    },
+    {
+      "epoch": 0.4149521877716604,
+      "grad_norm": 1.22348952293396,
+      "learning_rate": 1.7246713070378966e-05,
+      "loss": 0.9765,
+      "step": 179
+    },
+    {
+      "epoch": 0.41727035641842947,
+      "grad_norm": 1.0363351106643677,
+      "learning_rate": 1.7231245166279972e-05,
+      "loss": 1.0487,
+      "step": 180
+    },
+    {
+      "epoch": 0.4195885250651985,
+      "grad_norm": 0.8833026885986328,
+      "learning_rate": 1.7215777262180976e-05,
+      "loss": 1.0261,
+      "step": 181
+    },
+    {
+      "epoch": 0.42190669371196754,
+      "grad_norm": 0.8683452606201172,
+      "learning_rate": 1.7200309358081982e-05,
+      "loss": 1.0609,
+      "step": 182
+    },
+    {
+      "epoch": 0.4242248623587366,
+      "grad_norm": 0.8211922645568848,
+      "learning_rate": 1.718484145398299e-05,
+      "loss": 0.9942,
+      "step": 183
+    },
+    {
+      "epoch": 0.42654303100550567,
+      "grad_norm": 0.8936122059822083,
+      "learning_rate": 1.7169373549883992e-05,
+      "loss": 1.0591,
+      "step": 184
+    },
+    {
+      "epoch": 0.4288611996522747,
+      "grad_norm": 0.9455772042274475,
+      "learning_rate": 1.7153905645785e-05,
+      "loss": 1.0628,
+      "step": 185
+    },
+    {
+      "epoch": 0.43117936829904374,
+      "grad_norm": 1.0464543104171753,
+      "learning_rate": 1.7138437741686002e-05,
+      "loss": 0.9742,
+      "step": 186
+    },
+    {
+      "epoch": 0.43349753694581283,
+      "grad_norm": 1.4931954145431519,
+      "learning_rate": 1.712296983758701e-05,
+      "loss": 1.0534,
+      "step": 187
+    },
+    {
+      "epoch": 0.43581570559258187,
+      "grad_norm": 0.7873006463050842,
+      "learning_rate": 1.7107501933488012e-05,
+      "loss": 1.0499,
+      "step": 188
+    },
+    {
+      "epoch": 0.4381338742393509,
+      "grad_norm": 0.7451059222221375,
+      "learning_rate": 1.709203402938902e-05,
+      "loss": 1.0006,
+      "step": 189
+    },
+    {
+      "epoch": 0.44045204288611994,
+      "grad_norm": 0.8252111673355103,
+      "learning_rate": 1.7076566125290022e-05,
+      "loss": 1.057,
+      "step": 190
+    },
+    {
+      "epoch": 0.44277021153288904,
+      "grad_norm": 0.6444481015205383,
+      "learning_rate": 1.706109822119103e-05,
+      "loss": 1.0098,
+      "step": 191
+    },
+    {
+      "epoch": 0.44508838017965807,
+      "grad_norm": 0.8497568368911743,
+      "learning_rate": 1.7045630317092036e-05,
+      "loss": 1.0186,
+      "step": 192
+    },
+    {
+      "epoch": 0.4474065488264271,
+      "grad_norm": 1.0328199863433838,
+      "learning_rate": 1.7030162412993042e-05,
+      "loss": 1.0474,
+      "step": 193
+    },
+    {
+      "epoch": 0.4497247174731962,
+      "grad_norm": 0.7315878868103027,
+      "learning_rate": 1.7014694508894046e-05,
+      "loss": 0.9985,
+      "step": 194
+    },
+    {
+      "epoch": 0.45204288611996524,
+      "grad_norm": 1.0060752630233765,
+      "learning_rate": 1.6999226604795052e-05,
+      "loss": 1.0985,
+      "step": 195
+    },
+    {
+      "epoch": 0.4543610547667343,
+      "grad_norm": 0.8653793334960938,
+      "learning_rate": 1.698375870069606e-05,
+      "loss": 0.999,
+      "step": 196
+    },
+    {
+      "epoch": 0.4566792234135033,
+      "grad_norm": 1.0214215517044067,
+      "learning_rate": 1.6968290796597062e-05,
+      "loss": 1.0759,
+      "step": 197
+    },
+    {
+      "epoch": 0.4589973920602724,
+      "grad_norm": 0.7069177627563477,
+      "learning_rate": 1.695282289249807e-05,
+      "loss": 1.1569,
+      "step": 198
+    },
+    {
+      "epoch": 0.46131556070704144,
+      "grad_norm": 1.8065637350082397,
+      "learning_rate": 1.6937354988399072e-05,
+      "loss": 1.0398,
+      "step": 199
+    },
+    {
+      "epoch": 0.4636337293538105,
+      "grad_norm": 8.820870399475098,
+      "learning_rate": 1.692188708430008e-05,
+      "loss": 1.0801,
+      "step": 200
+    },
+    {
+      "epoch": 0.46595189800057957,
+      "grad_norm": 0.7957196235656738,
+      "learning_rate": 1.6906419180201082e-05,
+      "loss": 1.01,
+      "step": 201
+    },
+    {
+      "epoch": 0.4682700666473486,
+      "grad_norm": 0.9078807830810547,
+      "learning_rate": 1.689095127610209e-05,
+      "loss": 1.0068,
+      "step": 202
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 0.8792912364006042,
+      "learning_rate": 1.6875483372003096e-05,
+      "loss": 1.0169,
+      "step": 203
+    },
+    {
+      "epoch": 0.4729064039408867,
+      "grad_norm": 1.0289100408554077,
+      "learning_rate": 1.68600154679041e-05,
+      "loss": 0.9956,
+      "step": 204
+    },
+    {
+      "epoch": 0.47522457258765577,
+      "grad_norm": 2.788477897644043,
+      "learning_rate": 1.6844547563805106e-05,
+      "loss": 1.0839,
+      "step": 205
+    },
+    {
+      "epoch": 0.4775427412344248,
+      "grad_norm": 0.841396152973175,
+      "learning_rate": 1.6829079659706112e-05,
+      "loss": 1.0735,
+      "step": 206
+    },
+    {
+      "epoch": 0.47986090988119384,
+      "grad_norm": 1.1330881118774414,
+      "learning_rate": 1.681361175560712e-05,
+      "loss": 1.0486,
+      "step": 207
+    },
+    {
+      "epoch": 0.48217907852796293,
+      "grad_norm": 1.2185862064361572,
+      "learning_rate": 1.6798143851508122e-05,
+      "loss": 1.0962,
+      "step": 208
+    },
+    {
+      "epoch": 0.48449724717473197,
+      "grad_norm": 1.2216246128082275,
+      "learning_rate": 1.678267594740913e-05,
+      "loss": 1.0202,
+      "step": 209
+    },
+    {
+      "epoch": 0.486815415821501,
+      "grad_norm": 0.7822516560554504,
+      "learning_rate": 1.6767208043310132e-05,
+      "loss": 1.0278,
+      "step": 210
+    },
+    {
+      "epoch": 0.48913358446827004,
+      "grad_norm": 1.021760106086731,
+      "learning_rate": 1.675174013921114e-05,
+      "loss": 1.0626,
+      "step": 211
+    },
+    {
+      "epoch": 0.49145175311503914,
+      "grad_norm": 1.334328055381775,
+      "learning_rate": 1.6736272235112142e-05,
+      "loss": 1.1798,
+      "step": 212
+    },
+    {
+      "epoch": 0.4937699217618082,
+      "grad_norm": 0.7392475605010986,
+      "learning_rate": 1.672080433101315e-05,
+      "loss": 1.0713,
+      "step": 213
+    },
+    {
+      "epoch": 0.4960880904085772,
+      "grad_norm": 0.768805205821991,
+      "learning_rate": 1.6705336426914152e-05,
+      "loss": 0.9839,
+      "step": 214
+    },
+    {
+      "epoch": 0.4984062590553463,
+      "grad_norm": 0.7203591465950012,
+      "learning_rate": 1.668986852281516e-05,
+      "loss": 0.9926,
+      "step": 215
+    },
+    {
+      "epoch": 0.5007244277021153,
+      "grad_norm": 1.2835793495178223,
+      "learning_rate": 1.6674400618716166e-05,
+      "loss": 1.024,
+      "step": 216
+    },
+    {
+      "epoch": 0.5030425963488844,
+      "grad_norm": 0.8296486139297485,
+      "learning_rate": 1.6658932714617173e-05,
+      "loss": 1.0421,
+      "step": 217
+    },
+    {
+      "epoch": 0.5053607649956534,
+      "grad_norm": 0.8933680653572083,
+      "learning_rate": 1.6643464810518176e-05,
+      "loss": 0.9958,
+      "step": 218
+    },
+    {
+      "epoch": 0.5076789336424224,
+      "grad_norm": 0.6815921068191528,
+      "learning_rate": 1.6627996906419182e-05,
+      "loss": 1.0334,
+      "step": 219
+    },
+    {
+      "epoch": 0.5099971022891915,
+      "grad_norm": 0.795447051525116,
+      "learning_rate": 1.661252900232019e-05,
+      "loss": 1.0553,
+      "step": 220
+    },
+    {
+      "epoch": 0.5123152709359606,
+      "grad_norm": 1.1784237623214722,
+      "learning_rate": 1.6597061098221192e-05,
+      "loss": 0.9998,
+      "step": 221
+    },
+    {
+      "epoch": 0.5146334395827297,
+      "grad_norm": 0.9474261403083801,
+      "learning_rate": 1.65815931941222e-05,
+      "loss": 1.0704,
+      "step": 222
+    },
+    {
+      "epoch": 0.5169516082294987,
+      "grad_norm": 0.9175812602043152,
+      "learning_rate": 1.6566125290023202e-05,
+      "loss": 1.0779,
+      "step": 223
+    },
+    {
+      "epoch": 0.5192697768762677,
+      "grad_norm": 0.8006009459495544,
+      "learning_rate": 1.655065738592421e-05,
+      "loss": 1.024,
+      "step": 224
+    },
+    {
+      "epoch": 0.5215879455230368,
+      "grad_norm": 0.7539005875587463,
+      "learning_rate": 1.6535189481825212e-05,
+      "loss": 1.0301,
+      "step": 225
+    },
+    {
+      "epoch": 0.5239061141698058,
+      "grad_norm": 0.8373304009437561,
+      "learning_rate": 1.651972157772622e-05,
+      "loss": 1.0519,
+      "step": 226
+    },
+    {
+      "epoch": 0.5262242828165749,
+      "grad_norm": 0.7653727531433105,
+      "learning_rate": 1.6504253673627222e-05,
+      "loss": 0.9657,
+      "step": 227
+    },
+    {
+      "epoch": 0.528542451463344,
+      "grad_norm": 0.6552687287330627,
+      "learning_rate": 1.648878576952823e-05,
+      "loss": 1.0106,
+      "step": 228
+    },
+    {
+      "epoch": 0.530860620110113,
+      "grad_norm": 1.428830862045288,
+      "learning_rate": 1.6473317865429236e-05,
+      "loss": 1.0327,
+      "step": 229
+    },
+    {
+      "epoch": 0.5331787887568821,
+      "grad_norm": 1.0795942544937134,
+      "learning_rate": 1.6457849961330243e-05,
+      "loss": 1.0854,
+      "step": 230
+    },
+    {
+      "epoch": 0.5354969574036511,
+      "grad_norm": 0.5399507284164429,
+      "learning_rate": 1.644238205723125e-05,
+      "loss": 0.9932,
+      "step": 231
+    },
+    {
+      "epoch": 0.5378151260504201,
+      "grad_norm": 2.8251047134399414,
+      "learning_rate": 1.6426914153132253e-05,
+      "loss": 1.002,
+      "step": 232
+    },
+    {
+      "epoch": 0.5401332946971892,
+      "grad_norm": 0.7555001974105835,
+      "learning_rate": 1.641144624903326e-05,
+      "loss": 0.9898,
+      "step": 233
+    },
+    {
+      "epoch": 0.5424514633439582,
+      "grad_norm": 0.9090583324432373,
+      "learning_rate": 1.6395978344934263e-05,
+      "loss": 0.9762,
+      "step": 234
+    },
+    {
+      "epoch": 0.5447696319907274,
+      "grad_norm": 0.8169143199920654,
+      "learning_rate": 1.638051044083527e-05,
+      "loss": 1.0588,
+      "step": 235
+    },
+    {
+      "epoch": 0.5470878006374964,
+      "grad_norm": 0.7842413783073425,
+      "learning_rate": 1.6365042536736273e-05,
+      "loss": 0.9432,
+      "step": 236
+    },
+    {
+      "epoch": 0.5494059692842654,
+      "grad_norm": 2.24771785736084,
+      "learning_rate": 1.634957463263728e-05,
+      "loss": 1.0636,
+      "step": 237
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 0.9846341013908386,
+      "learning_rate": 1.6334106728538283e-05,
+      "loss": 0.9743,
+      "step": 238
+    },
+    {
+      "epoch": 0.5540423065778035,
+      "grad_norm": 0.7598584294319153,
+      "learning_rate": 1.631863882443929e-05,
+      "loss": 1.0192,
+      "step": 239
+    },
+    {
+      "epoch": 0.5563604752245725,
+      "grad_norm": 1.200215458869934,
+      "learning_rate": 1.6303170920340296e-05,
+      "loss": 1.1051,
+      "step": 240
+    },
+    {
+      "epoch": 0.5586786438713416,
+      "grad_norm": 0.8878689408302307,
+      "learning_rate": 1.62877030162413e-05,
+      "loss": 1.0622,
+      "step": 241
+    },
+    {
+      "epoch": 0.5609968125181107,
+      "grad_norm": 1.18966543674469,
+      "learning_rate": 1.6272235112142306e-05,
+      "loss": 1.0396,
+      "step": 242
+    },
+    {
+      "epoch": 0.5633149811648798,
+      "grad_norm": 0.7161230444908142,
+      "learning_rate": 1.6256767208043313e-05,
+      "loss": 1.0611,
+      "step": 243
+    },
+    {
+      "epoch": 0.5656331498116488,
+      "grad_norm": 0.806091845035553,
+      "learning_rate": 1.624129930394432e-05,
+      "loss": 1.0862,
+      "step": 244
+    },
+    {
+      "epoch": 0.5679513184584178,
+      "grad_norm": 0.9060055613517761,
+      "learning_rate": 1.6225831399845323e-05,
+      "loss": 0.9447,
+      "step": 245
+    },
+    {
+      "epoch": 0.5702694871051869,
+      "grad_norm": 0.6209578514099121,
+      "learning_rate": 1.621036349574633e-05,
+      "loss": 1.0291,
+      "step": 246
+    },
+    {
+      "epoch": 0.5725876557519559,
+      "grad_norm": 0.8875139951705933,
+      "learning_rate": 1.6194895591647333e-05,
+      "loss": 0.9596,
+      "step": 247
+    },
+    {
+      "epoch": 0.5749058243987251,
+      "grad_norm": 0.8894098997116089,
+      "learning_rate": 1.617942768754834e-05,
+      "loss": 1.0363,
+      "step": 248
+    },
+    {
+      "epoch": 0.5772239930454941,
+      "grad_norm": 0.8011065125465393,
+      "learning_rate": 1.6163959783449343e-05,
+      "loss": 0.9735,
+      "step": 249
+    },
+    {
+      "epoch": 0.5795421616922631,
+      "grad_norm": 1.0448037385940552,
+      "learning_rate": 1.614849187935035e-05,
+      "loss": 1.0154,
+      "step": 250
+    },
+    {
+      "epoch": 0.5818603303390322,
+      "grad_norm": 0.7367164492607117,
+      "learning_rate": 1.6133023975251353e-05,
+      "loss": 1.0509,
+      "step": 251
+    },
+    {
+      "epoch": 0.5841784989858012,
+      "grad_norm": 0.8820902705192566,
+      "learning_rate": 1.611755607115236e-05,
+      "loss": 1.074,
+      "step": 252
+    },
+    {
+      "epoch": 0.5864966676325702,
+      "grad_norm": 0.8512645363807678,
+      "learning_rate": 1.6102088167053366e-05,
+      "loss": 1.0478,
+      "step": 253
+    },
+    {
+      "epoch": 0.5888148362793393,
+      "grad_norm": 0.8832964897155762,
+      "learning_rate": 1.6086620262954373e-05,
+      "loss": 1.0502,
+      "step": 254
+    },
+    {
+      "epoch": 0.5911330049261084,
+      "grad_norm": 0.7311517596244812,
+      "learning_rate": 1.6071152358855376e-05,
+      "loss": 0.9904,
+      "step": 255
+    },
+    {
+      "epoch": 0.5934511735728775,
+      "grad_norm": 0.9509069919586182,
+      "learning_rate": 1.6055684454756383e-05,
+      "loss": 1.0628,
+      "step": 256
+    },
+    {
+      "epoch": 0.5957693422196465,
+      "grad_norm": 0.5056537389755249,
+      "learning_rate": 1.604021655065739e-05,
+      "loss": 0.9579,
+      "step": 257
+    },
+    {
+      "epoch": 0.5980875108664155,
+      "grad_norm": 0.6654573082923889,
+      "learning_rate": 1.6024748646558393e-05,
+      "loss": 0.9752,
+      "step": 258
+    },
+    {
+      "epoch": 0.6004056795131846,
+      "grad_norm": 0.7242197394371033,
+      "learning_rate": 1.60092807424594e-05,
+      "loss": 1.0986,
+      "step": 259
+    },
+    {
+      "epoch": 0.6027238481599536,
+      "grad_norm": 0.8016011118888855,
+      "learning_rate": 1.5993812838360403e-05,
+      "loss": 1.0193,
+      "step": 260
+    },
+    {
+      "epoch": 0.6050420168067226,
+      "grad_norm": 0.8806138038635254,
+      "learning_rate": 1.597834493426141e-05,
+      "loss": 1.0192,
+      "step": 261
+    },
+    {
+      "epoch": 0.6073601854534918,
+      "grad_norm": 1.1127492189407349,
+      "learning_rate": 1.5962877030162413e-05,
+      "loss": 1.1279,
+      "step": 262
+    },
+    {
+      "epoch": 0.6096783541002608,
+      "grad_norm": 0.8649683594703674,
+      "learning_rate": 1.594740912606342e-05,
+      "loss": 1.0059,
+      "step": 263
+    },
+    {
+      "epoch": 0.6119965227470299,
+      "grad_norm": 0.7879909873008728,
+      "learning_rate": 1.5931941221964423e-05,
+      "loss": 1.0077,
+      "step": 264
+    },
+    {
+      "epoch": 0.6143146913937989,
+      "grad_norm": 0.8802973031997681,
+      "learning_rate": 1.591647331786543e-05,
+      "loss": 1.0484,
+      "step": 265
+    },
+    {
+      "epoch": 0.6166328600405679,
+      "grad_norm": 1.8282607793807983,
+      "learning_rate": 1.5901005413766436e-05,
+      "loss": 0.9435,
+      "step": 266
+    },
+    {
+      "epoch": 0.618951028687337,
+      "grad_norm": 0.643280565738678,
+      "learning_rate": 1.5885537509667443e-05,
+      "loss": 0.948,
+      "step": 267
+    },
+    {
+      "epoch": 0.621269197334106,
+      "grad_norm": 0.727376401424408,
+      "learning_rate": 1.587006960556845e-05,
+      "loss": 0.9687,
+      "step": 268
+    },
+    {
+      "epoch": 0.6235873659808752,
+      "grad_norm": 0.9891621470451355,
+      "learning_rate": 1.5854601701469453e-05,
+      "loss": 0.995,
+      "step": 269
+    },
+    {
+      "epoch": 0.6259055346276442,
+      "grad_norm": 1.4208780527114868,
+      "learning_rate": 1.583913379737046e-05,
+      "loss": 1.107,
+      "step": 270
+    },
+    {
+      "epoch": 0.6282237032744132,
+      "grad_norm": 0.8574293851852417,
+      "learning_rate": 1.5823665893271463e-05,
+      "loss": 1.0016,
+      "step": 271
+    },
+    {
+      "epoch": 0.6305418719211823,
+      "grad_norm": 1.4257714748382568,
+      "learning_rate": 1.580819798917247e-05,
+      "loss": 1.3383,
+      "step": 272
+    },
+    {
+      "epoch": 0.6328600405679513,
+      "grad_norm": 0.7138167023658752,
+      "learning_rate": 1.5792730085073473e-05,
+      "loss": 1.0534,
+      "step": 273
+    },
+    {
+      "epoch": 0.6351782092147203,
+      "grad_norm": 1.5973683595657349,
+      "learning_rate": 1.577726218097448e-05,
+      "loss": 0.9784,
+      "step": 274
+    },
+    {
+      "epoch": 0.6374963778614894,
+      "grad_norm": 0.6794442534446716,
+      "learning_rate": 1.5761794276875483e-05,
+      "loss": 0.9809,
+      "step": 275
+    },
+    {
+      "epoch": 0.6398145465082585,
+      "grad_norm": 0.7616905570030212,
+      "learning_rate": 1.574632637277649e-05,
+      "loss": 0.9981,
+      "step": 276
+    },
+    {
+      "epoch": 0.6421327151550276,
+      "grad_norm": 1.709405541419983,
+      "learning_rate": 1.5730858468677496e-05,
+      "loss": 1.0105,
+      "step": 277
+    },
+    {
+      "epoch": 0.6444508838017966,
+      "grad_norm": 0.6796721816062927,
+      "learning_rate": 1.57153905645785e-05,
+      "loss": 0.9856,
+      "step": 278
+    },
+    {
+      "epoch": 0.6467690524485656,
+      "grad_norm": 0.7686854600906372,
+      "learning_rate": 1.5699922660479506e-05,
+      "loss": 1.0534,
+      "step": 279
+    },
+    {
+      "epoch": 0.6490872210953347,
+      "grad_norm": 1.0257889032363892,
+      "learning_rate": 1.5684454756380513e-05,
+      "loss": 0.9869,
+      "step": 280
+    },
+    {
+      "epoch": 0.6514053897421037,
+      "grad_norm": 0.7100695371627808,
+      "learning_rate": 1.566898685228152e-05,
+      "loss": 1.0433,
+      "step": 281
+    },
+    {
+      "epoch": 0.6537235583888727,
+      "grad_norm": 0.7201927900314331,
+      "learning_rate": 1.5653518948182523e-05,
+      "loss": 0.9854,
+      "step": 282
+    },
+    {
+      "epoch": 0.6560417270356419,
+      "grad_norm": 1.5743852853775024,
+      "learning_rate": 1.563805104408353e-05,
+      "loss": 1.0076,
+      "step": 283
+    },
+    {
+      "epoch": 0.6583598956824109,
+      "grad_norm": 0.7456634640693665,
+      "learning_rate": 1.5622583139984533e-05,
+      "loss": 1.0214,
+      "step": 284
+    },
+    {
+      "epoch": 0.66067806432918,
+      "grad_norm": 0.6395049691200256,
+      "learning_rate": 1.560711523588554e-05,
+      "loss": 1.0321,
+      "step": 285
+    },
+    {
+      "epoch": 0.662996232975949,
+      "grad_norm": 0.9406479001045227,
+      "learning_rate": 1.5591647331786543e-05,
+      "loss": 1.0387,
+      "step": 286
+    },
+    {
+      "epoch": 0.665314401622718,
+      "grad_norm": 0.668521523475647,
+      "learning_rate": 1.557617942768755e-05,
+      "loss": 0.9458,
+      "step": 287
+    },
+    {
+      "epoch": 0.6676325702694871,
+      "grad_norm": 0.8241714239120483,
+      "learning_rate": 1.5560711523588553e-05,
+      "loss": 0.994,
+      "step": 288
+    },
+    {
+      "epoch": 0.6699507389162561,
+      "grad_norm": 0.7906151413917542,
+      "learning_rate": 1.554524361948956e-05,
+      "loss": 1.0127,
+      "step": 289
+    },
+    {
+      "epoch": 0.6722689075630253,
+      "grad_norm": 3.7441999912261963,
+      "learning_rate": 1.5529775715390566e-05,
+      "loss": 1.1235,
+      "step": 290
+    },
+    {
+      "epoch": 0.6745870762097943,
+      "grad_norm": 0.7488934397697449,
+      "learning_rate": 1.5514307811291573e-05,
+      "loss": 1.0132,
+      "step": 291
+    },
+    {
+      "epoch": 0.6769052448565633,
+      "grad_norm": 0.6223219037055969,
+      "learning_rate": 1.5498839907192576e-05,
+      "loss": 0.9831,
+      "step": 292
+    },
+    {
+      "epoch": 0.6792234135033324,
+      "grad_norm": 0.9072495698928833,
+      "learning_rate": 1.5483372003093583e-05,
+      "loss": 0.9978,
+      "step": 293
+    },
+    {
+      "epoch": 0.6815415821501014,
+      "grad_norm": 1.0984013080596924,
+      "learning_rate": 1.546790409899459e-05,
+      "loss": 0.9942,
+      "step": 294
+    },
+    {
+      "epoch": 0.6838597507968704,
+      "grad_norm": 0.8855274319648743,
+      "learning_rate": 1.5452436194895593e-05,
+      "loss": 1.0135,
+      "step": 295
+    },
+    {
+      "epoch": 0.6861779194436395,
+      "grad_norm": 0.7710789442062378,
+      "learning_rate": 1.54369682907966e-05,
+      "loss": 1.034,
+      "step": 296
+    },
+    {
+      "epoch": 0.6884960880904086,
+      "grad_norm": 19.71253204345703,
+      "learning_rate": 1.5421500386697603e-05,
+      "loss": 0.9598,
+      "step": 297
+    },
+    {
+      "epoch": 0.6908142567371777,
+      "grad_norm": 0.930057942867279,
+      "learning_rate": 1.540603248259861e-05,
+      "loss": 0.9755,
+      "step": 298
+    },
+    {
+      "epoch": 0.6931324253839467,
+      "grad_norm": 1.0376302003860474,
+      "learning_rate": 1.5390564578499613e-05,
+      "loss": 1.0324,
+      "step": 299
+    },
+    {
+      "epoch": 0.6954505940307157,
+      "grad_norm": 0.8160769939422607,
+      "learning_rate": 1.537509667440062e-05,
+      "loss": 1.0805,
+      "step": 300
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1293,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0925139234599731e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aaf0b1b95273dbdf80afb35d7fc535ee7d5ea84d998c877b4bc52a6963322dcc
+size 7352