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

added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<image_soft_token>": 262144
+}

config.json

@@ -0,0 +1,37 @@
+{
+  "architectures": [
+    "Gemma3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "attn_logit_softcapping": null,
+  "bos_token_id": 2,
+  "cache_implementation": "hybrid",
+  "eos_token_id": 106,
+  "final_logit_softcapping": null,
+  "head_dim": 256,
+  "hidden_activation": "gelu_pytorch_tanh",
+  "hidden_size": 2560,
+  "initializer_range": 0.02,
+  "intermediate_size": 10240,
+  "max_position_embeddings": 131072,
+  "model_type": "gemma3_text",
+  "num_attention_heads": 8,
+  "num_hidden_layers": 34,
+  "num_key_value_heads": 4,
+  "pad_token_id": 0,
+  "query_pre_attn_scalar": 256,
+  "rms_norm_eps": 1e-06,
+  "rope_local_base_freq": 10000.0,
+  "rope_scaling": {
+    "factor": 8.0,
+    "rope_type": "linear"
+  },
+  "rope_theta": 1000000.0,
+  "sliding_window": 1024,
+  "sliding_window_pattern": 6,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.0.dev0",
+  "use_cache": false,
+  "vocab_size": 262208
+}

generation_config.json

@@ -0,0 +1,13 @@
+{
+  "bos_token_id": 2,
+  "cache_implementation": "hybrid",
+  "do_sample": true,
+  "eos_token_id": [
+    1,
+    106
+  ],
+  "pad_token_id": 0,
+  "top_k": 64,
+  "top_p": 0.95,
+  "transformers_version": "4.51.0.dev0"
+}

model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b0f96acb87efa03434f396b0198d5a8f61dcc0501a9e5cee3d5754a96f85f80
+size 4960531344

model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6af45d8a5d17c6037a736e314cd6496f115953ee956f5bab942b847b71de0580
+size 2800046672

model.safetensors.index.json

@@ -0,0 +1,451 @@
+{
+  "metadata": {
+    "total_size": 7760526336
+  },
+  "weight_map": {
+    "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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.pre_feedforward_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-00002-of-00002.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.19.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.19.pre_feedforward_layernorm.weight": "model-00002-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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.pre_feedforward_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-00002-of-00002.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00002-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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.pre_feedforward_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.30.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00002-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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.pre_feedforward_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.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.pre_feedforward_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.norm.weight": "model-00002-of-00002.safetensors"
+  }
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b604a84803959f44a0e5209aaa17164896a151670ffe5468ae3309398f0bc53
+size 936

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67fcbf37e2292d5bb13bc3717341ad7a0a6a32c620d147778e6bb8c5a9d81ab4
+size 14244

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:516d67c75e20f1511f3dbc7fa9eb52f47cd711a1ee777d45a96033b26ea5156d
+size 1064

special_tokens_map.json

@@ -0,0 +1,33 @@
+{
+  "boi_token": "<start_of_image>",
+  "bos_token": {
+    "content": "<bos>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eoi_token": "<end_of_image>",
+  "eos_token": {
+    "content": "<end_of_turn>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "image_token": "<image_soft_token>",
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "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 sha256:4667f2089529e8e7657cfb6d1c19910ae71ff5f28aa7ab2ff2763330affad795
+size 33384568

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c
+size 4689074

trainer_state.json

@@ -0,0 +1,3072 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 434,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002304147465437788,
+      "grad_norm": 0.0,
+      "learning_rate": 9.998846730480915e-06,
+      "loss": 3.6297,
+      "step": 1
+    },
+    {
+      "epoch": 0.004608294930875576,
+      "grad_norm": 0.0,
+      "learning_rate": 9.997691064419303e-06,
+      "loss": 3.7821,
+      "step": 2
+    },
+    {
+      "epoch": 0.0069124423963133645,
+      "grad_norm": 0.0,
+      "learning_rate": 9.996532994337224e-06,
+      "loss": 3.0545,
+      "step": 3
+    },
+    {
+      "epoch": 0.009216589861751152,
+      "grad_norm": 0.0,
+      "learning_rate": 9.99537251272559e-06,
+      "loss": 3.3776,
+      "step": 4
+    },
+    {
+      "epoch": 0.01152073732718894,
+      "grad_norm": 0.0,
+      "learning_rate": 9.994209612044008e-06,
+      "loss": 2.6629,
+      "step": 5
+    },
+    {
+      "epoch": 0.013824884792626729,
+      "grad_norm": 0.0,
+      "learning_rate": 9.993044284720612e-06,
+      "loss": 2.9764,
+      "step": 6
+    },
+    {
+      "epoch": 0.016129032258064516,
+      "grad_norm": 0.0,
+      "learning_rate": 9.99187652315191e-06,
+      "loss": 3.0443,
+      "step": 7
+    },
+    {
+      "epoch": 0.018433179723502304,
+      "grad_norm": 0.0,
+      "learning_rate": 9.990706319702603e-06,
+      "loss": 2.9777,
+      "step": 8
+    },
+    {
+      "epoch": 0.020737327188940093,
+      "grad_norm": 0.0,
+      "learning_rate": 9.989533666705432e-06,
+      "loss": 2.6634,
+      "step": 9
+    },
+    {
+      "epoch": 0.02304147465437788,
+      "grad_norm": 0.0,
+      "learning_rate": 9.988358556461003e-06,
+      "loss": 2.5688,
+      "step": 10
+    },
+    {
+      "epoch": 0.02534562211981567,
+      "grad_norm": 0.0,
+      "learning_rate": 9.98718098123762e-06,
+      "loss": 2.7999,
+      "step": 11
+    },
+    {
+      "epoch": 0.027649769585253458,
+      "grad_norm": 0.0,
+      "learning_rate": 9.986000933271115e-06,
+      "loss": 2.8395,
+      "step": 12
+    },
+    {
+      "epoch": 0.029953917050691243,
+      "grad_norm": 0.0,
+      "learning_rate": 9.984818404764686e-06,
+      "loss": 2.9181,
+      "step": 13
+    },
+    {
+      "epoch": 0.03225806451612903,
+      "grad_norm": 0.0,
+      "learning_rate": 9.983633387888708e-06,
+      "loss": 2.3633,
+      "step": 14
+    },
+    {
+      "epoch": 0.03456221198156682,
+      "grad_norm": 0.0,
+      "learning_rate": 9.982445874780575e-06,
+      "loss": 2.6805,
+      "step": 15
+    },
+    {
+      "epoch": 0.03686635944700461,
+      "grad_norm": 0.0,
+      "learning_rate": 9.981255857544517e-06,
+      "loss": 3.2891,
+      "step": 16
+    },
+    {
+      "epoch": 0.03917050691244239,
+      "grad_norm": 0.0,
+      "learning_rate": 9.980063328251437e-06,
+      "loss": 2.3438,
+      "step": 17
+    },
+    {
+      "epoch": 0.041474654377880185,
+      "grad_norm": 0.0,
+      "learning_rate": 9.97886827893872e-06,
+      "loss": 3.0867,
+      "step": 18
+    },
+    {
+      "epoch": 0.04377880184331797,
+      "grad_norm": 0.0,
+      "learning_rate": 9.97767070161006e-06,
+      "loss": 3.0397,
+      "step": 19
+    },
+    {
+      "epoch": 0.04608294930875576,
+      "grad_norm": 0.0,
+      "learning_rate": 9.976470588235294e-06,
+      "loss": 2.675,
+      "step": 20
+    },
+    {
+      "epoch": 0.04838709677419355,
+      "grad_norm": 0.0,
+      "learning_rate": 9.975267930750206e-06,
+      "loss": 2.588,
+      "step": 21
+    },
+    {
+      "epoch": 0.05069124423963134,
+      "grad_norm": 0.0,
+      "learning_rate": 9.974062721056355e-06,
+      "loss": 2.5917,
+      "step": 22
+    },
+    {
+      "epoch": 0.052995391705069124,
+      "grad_norm": 0.0,
+      "learning_rate": 9.97285495102089e-06,
+      "loss": 2.6421,
+      "step": 23
+    },
+    {
+      "epoch": 0.055299539170506916,
+      "grad_norm": 0.0,
+      "learning_rate": 9.971644612476371e-06,
+      "loss": 2.8325,
+      "step": 24
+    },
+    {
+      "epoch": 0.0576036866359447,
+      "grad_norm": 0.0,
+      "learning_rate": 9.970431697220581e-06,
+      "loss": 2.7928,
+      "step": 25
+    },
+    {
+      "epoch": 0.059907834101382486,
+      "grad_norm": 0.0,
+      "learning_rate": 9.96921619701634e-06,
+      "loss": 2.7043,
+      "step": 26
+    },
+    {
+      "epoch": 0.06221198156682028,
+      "grad_norm": 0.0,
+      "learning_rate": 9.967998103591326e-06,
+      "loss": 2.5606,
+      "step": 27
+    },
+    {
+      "epoch": 0.06451612903225806,
+      "grad_norm": 0.0,
+      "learning_rate": 9.966777408637875e-06,
+      "loss": 2.5944,
+      "step": 28
+    },
+    {
+      "epoch": 0.06682027649769585,
+      "grad_norm": 0.0,
+      "learning_rate": 9.965554103812804e-06,
+      "loss": 2.0101,
+      "step": 29
+    },
+    {
+      "epoch": 0.06912442396313365,
+      "grad_norm": 0.0,
+      "learning_rate": 9.964328180737219e-06,
+      "loss": 2.669,
+      "step": 30
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 0.0,
+      "learning_rate": 9.963099630996312e-06,
+      "loss": 2.5637,
+      "step": 31
+    },
+    {
+      "epoch": 0.07373271889400922,
+      "grad_norm": 0.0,
+      "learning_rate": 9.961868446139181e-06,
+      "loss": 2.8749,
+      "step": 32
+    },
+    {
+      "epoch": 0.07603686635944701,
+      "grad_norm": 0.0,
+      "learning_rate": 9.960634617678637e-06,
+      "loss": 2.8018,
+      "step": 33
+    },
+    {
+      "epoch": 0.07834101382488479,
+      "grad_norm": 0.0,
+      "learning_rate": 9.959398137090996e-06,
+      "loss": 2.5725,
+      "step": 34
+    },
+    {
+      "epoch": 0.08064516129032258,
+      "grad_norm": 0.0,
+      "learning_rate": 9.958158995815901e-06,
+      "loss": 2.8989,
+      "step": 35
+    },
+    {
+      "epoch": 0.08294930875576037,
+      "grad_norm": 0.0,
+      "learning_rate": 9.956917185256105e-06,
+      "loss": 2.4909,
+      "step": 36
+    },
+    {
+      "epoch": 0.08525345622119816,
+      "grad_norm": 0.0,
+      "learning_rate": 9.955672696777287e-06,
+      "loss": 2.5061,
+      "step": 37
+    },
+    {
+      "epoch": 0.08755760368663594,
+      "grad_norm": 0.0,
+      "learning_rate": 9.954425521707844e-06,
+      "loss": 2.3851,
+      "step": 38
+    },
+    {
+      "epoch": 0.08986175115207373,
+      "grad_norm": 0.0,
+      "learning_rate": 9.953175651338697e-06,
+      "loss": 3.0129,
+      "step": 39
+    },
+    {
+      "epoch": 0.09216589861751152,
+      "grad_norm": 0.0,
+      "learning_rate": 9.951923076923079e-06,
+      "loss": 2.8212,
+      "step": 40
+    },
+    {
+      "epoch": 0.0944700460829493,
+      "grad_norm": 0.0,
+      "learning_rate": 9.950667789676334e-06,
+      "loss": 2.952,
+      "step": 41
+    },
+    {
+      "epoch": 0.0967741935483871,
+      "grad_norm": 0.0,
+      "learning_rate": 9.949409780775717e-06,
+      "loss": 2.2857,
+      "step": 42
+    },
+    {
+      "epoch": 0.09907834101382489,
+      "grad_norm": 0.0,
+      "learning_rate": 9.948149041360185e-06,
+      "loss": 2.2268,
+      "step": 43
+    },
+    {
+      "epoch": 0.10138248847926268,
+      "grad_norm": 0.0,
+      "learning_rate": 9.94688556253018e-06,
+      "loss": 2.7615,
+      "step": 44
+    },
+    {
+      "epoch": 0.10368663594470046,
+      "grad_norm": 0.0,
+      "learning_rate": 9.945619335347432e-06,
+      "loss": 3.1013,
+      "step": 45
+    },
+    {
+      "epoch": 0.10599078341013825,
+      "grad_norm": 0.0,
+      "learning_rate": 9.944350350834746e-06,
+      "loss": 3.0616,
+      "step": 46
+    },
+    {
+      "epoch": 0.10829493087557604,
+      "grad_norm": 0.0,
+      "learning_rate": 9.943078599975779e-06,
+      "loss": 2.2084,
+      "step": 47
+    },
+    {
+      "epoch": 0.11059907834101383,
+      "grad_norm": 0.0,
+      "learning_rate": 9.941804073714841e-06,
+      "loss": 2.1988,
+      "step": 48
+    },
+    {
+      "epoch": 0.11290322580645161,
+      "grad_norm": 0.0,
+      "learning_rate": 9.94052676295667e-06,
+      "loss": 2.5741,
+      "step": 49
+    },
+    {
+      "epoch": 0.1152073732718894,
+      "grad_norm": 0.0,
+      "learning_rate": 9.939246658566223e-06,
+      "loss": 2.4302,
+      "step": 50
+    },
+    {
+      "epoch": 0.1175115207373272,
+      "grad_norm": 0.0,
+      "learning_rate": 9.937963751368447e-06,
+      "loss": 2.5109,
+      "step": 51
+    },
+    {
+      "epoch": 0.11981566820276497,
+      "grad_norm": 0.0,
+      "learning_rate": 9.936678032148077e-06,
+      "loss": 2.2233,
+      "step": 52
+    },
+    {
+      "epoch": 0.12211981566820276,
+      "grad_norm": 0.0,
+      "learning_rate": 9.935389491649398e-06,
+      "loss": 3.1745,
+      "step": 53
+    },
+    {
+      "epoch": 0.12442396313364056,
+      "grad_norm": 0.0,
+      "learning_rate": 9.934098120576033e-06,
+      "loss": 2.6837,
+      "step": 54
+    },
+    {
+      "epoch": 0.12672811059907835,
+      "grad_norm": 0.0,
+      "learning_rate": 9.932803909590715e-06,
+      "loss": 2.5496,
+      "step": 55
+    },
+    {
+      "epoch": 0.12903225806451613,
+      "grad_norm": 0.0,
+      "learning_rate": 9.931506849315069e-06,
+      "loss": 2.6769,
+      "step": 56
+    },
+    {
+      "epoch": 0.1313364055299539,
+      "grad_norm": 0.0,
+      "learning_rate": 9.930206930329376e-06,
+      "loss": 2.4358,
+      "step": 57
+    },
+    {
+      "epoch": 0.1336405529953917,
+      "grad_norm": 0.0,
+      "learning_rate": 9.928904143172348e-06,
+      "loss": 2.6133,
+      "step": 58
+    },
+    {
+      "epoch": 0.1359447004608295,
+      "grad_norm": 0.0,
+      "learning_rate": 9.927598478340902e-06,
+      "loss": 2.6126,
+      "step": 59
+    },
+    {
+      "epoch": 0.1382488479262673,
+      "grad_norm": 0.0,
+      "learning_rate": 9.926289926289927e-06,
+      "loss": 2.3671,
+      "step": 60
+    },
+    {
+      "epoch": 0.14055299539170507,
+      "grad_norm": 0.0,
+      "learning_rate": 9.92497847743205e-06,
+      "loss": 2.3528,
+      "step": 61
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.0,
+      "learning_rate": 9.923664122137405e-06,
+      "loss": 2.894,
+      "step": 62
+    },
+    {
+      "epoch": 0.14516129032258066,
+      "grad_norm": 0.0,
+      "learning_rate": 9.922346850733393e-06,
+      "loss": 2.944,
+      "step": 63
+    },
+    {
+      "epoch": 0.14746543778801843,
+      "grad_norm": 0.0,
+      "learning_rate": 9.921026653504443e-06,
+      "loss": 2.495,
+      "step": 64
+    },
+    {
+      "epoch": 0.1497695852534562,
+      "grad_norm": 0.0,
+      "learning_rate": 9.919703520691786e-06,
+      "loss": 2.2919,
+      "step": 65
+    },
+    {
+      "epoch": 0.15207373271889402,
+      "grad_norm": 0.0,
+      "learning_rate": 9.918377442493198e-06,
+      "loss": 2.6164,
+      "step": 66
+    },
+    {
+      "epoch": 0.1543778801843318,
+      "grad_norm": 0.0,
+      "learning_rate": 9.917048409062771e-06,
+      "loss": 2.5487,
+      "step": 67
+    },
+    {
+      "epoch": 0.15668202764976957,
+      "grad_norm": 0.0,
+      "learning_rate": 9.91571641051066e-06,
+      "loss": 2.6602,
+      "step": 68
+    },
+    {
+      "epoch": 0.15898617511520738,
+      "grad_norm": 0.0,
+      "learning_rate": 9.914381436902843e-06,
+      "loss": 2.5056,
+      "step": 69
+    },
+    {
+      "epoch": 0.16129032258064516,
+      "grad_norm": 0.0,
+      "learning_rate": 9.913043478260871e-06,
+      "loss": 2.8503,
+      "step": 70
+    },
+    {
+      "epoch": 0.16359447004608296,
+      "grad_norm": 0.0,
+      "learning_rate": 9.911702524561623e-06,
+      "loss": 2.6178,
+      "step": 71
+    },
+    {
+      "epoch": 0.16589861751152074,
+      "grad_norm": 0.0,
+      "learning_rate": 9.910358565737052e-06,
+      "loss": 2.7173,
+      "step": 72
+    },
+    {
+      "epoch": 0.16820276497695852,
+      "grad_norm": 0.0,
+      "learning_rate": 9.909011591673938e-06,
+      "loss": 2.4166,
+      "step": 73
+    },
+    {
+      "epoch": 0.17050691244239632,
+      "grad_norm": 0.0,
+      "learning_rate": 9.907661592213627e-06,
+      "loss": 2.5658,
+      "step": 74
+    },
+    {
+      "epoch": 0.1728110599078341,
+      "grad_norm": 0.0,
+      "learning_rate": 9.906308557151781e-06,
+      "loss": 2.3656,
+      "step": 75
+    },
+    {
+      "epoch": 0.17511520737327188,
+      "grad_norm": 0.0,
+      "learning_rate": 9.90495247623812e-06,
+      "loss": 2.4611,
+      "step": 76
+    },
+    {
+      "epoch": 0.1774193548387097,
+      "grad_norm": 0.0,
+      "learning_rate": 9.903593339176162e-06,
+      "loss": 2.2937,
+      "step": 77
+    },
+    {
+      "epoch": 0.17972350230414746,
+      "grad_norm": 0.0,
+      "learning_rate": 9.902231135622965e-06,
+      "loss": 2.172,
+      "step": 78
+    },
+    {
+      "epoch": 0.18202764976958524,
+      "grad_norm": 0.0,
+      "learning_rate": 9.900865855188858e-06,
+      "loss": 2.4489,
+      "step": 79
+    },
+    {
+      "epoch": 0.18433179723502305,
+      "grad_norm": 0.0,
+      "learning_rate": 9.899497487437187e-06,
+      "loss": 2.7153,
+      "step": 80
+    },
+    {
+      "epoch": 0.18663594470046083,
+      "grad_norm": 0.0,
+      "learning_rate": 9.89812602188404e-06,
+      "loss": 2.7078,
+      "step": 81
+    },
+    {
+      "epoch": 0.1889400921658986,
+      "grad_norm": 0.0,
+      "learning_rate": 9.896751447997987e-06,
+      "loss": 2.1448,
+      "step": 82
+    },
+    {
+      "epoch": 0.1912442396313364,
+      "grad_norm": 0.0,
+      "learning_rate": 9.8953737551998e-06,
+      "loss": 2.4494,
+      "step": 83
+    },
+    {
+      "epoch": 0.1935483870967742,
+      "grad_norm": 0.0,
+      "learning_rate": 9.893992932862191e-06,
+      "loss": 2.5461,
+      "step": 84
+    },
+    {
+      "epoch": 0.195852534562212,
+      "grad_norm": 0.0,
+      "learning_rate": 9.89260897030954e-06,
+      "loss": 2.219,
+      "step": 85
+    },
+    {
+      "epoch": 0.19815668202764977,
+      "grad_norm": 0.0,
+      "learning_rate": 9.891221856817608e-06,
+      "loss": 2.6866,
+      "step": 86
+    },
+    {
+      "epoch": 0.20046082949308755,
+      "grad_norm": 0.0,
+      "learning_rate": 9.889831581613272e-06,
+      "loss": 2.2086,
+      "step": 87
+    },
+    {
+      "epoch": 0.20276497695852536,
+      "grad_norm": 0.0,
+      "learning_rate": 9.88843813387424e-06,
+      "loss": 2.658,
+      "step": 88
+    },
+    {
+      "epoch": 0.20506912442396313,
+      "grad_norm": 0.0,
+      "learning_rate": 9.887041502728773e-06,
+      "loss": 2.2961,
+      "step": 89
+    },
+    {
+      "epoch": 0.2073732718894009,
+      "grad_norm": 0.0,
+      "learning_rate": 9.885641677255401e-06,
+      "loss": 2.5561,
+      "step": 90
+    },
+    {
+      "epoch": 0.20967741935483872,
+      "grad_norm": 0.0,
+      "learning_rate": 9.884238646482637e-06,
+      "loss": 2.5409,
+      "step": 91
+    },
+    {
+      "epoch": 0.2119815668202765,
+      "grad_norm": 0.0,
+      "learning_rate": 9.882832399388692e-06,
+      "loss": 2.2584,
+      "step": 92
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 0.0,
+      "learning_rate": 9.881422924901188e-06,
+      "loss": 2.5798,
+      "step": 93
+    },
+    {
+      "epoch": 0.21658986175115208,
+      "grad_norm": 0.0,
+      "learning_rate": 9.88001021189686e-06,
+      "loss": 2.6547,
+      "step": 94
+    },
+    {
+      "epoch": 0.21889400921658986,
+      "grad_norm": 0.0,
+      "learning_rate": 9.878594249201278e-06,
+      "loss": 1.9719,
+      "step": 95
+    },
+    {
+      "epoch": 0.22119815668202766,
+      "grad_norm": 0.0,
+      "learning_rate": 9.877175025588538e-06,
+      "loss": 2.6163,
+      "step": 96
+    },
+    {
+      "epoch": 0.22350230414746544,
+      "grad_norm": 0.0,
+      "learning_rate": 9.875752529780967e-06,
+      "loss": 2.5005,
+      "step": 97
+    },
+    {
+      "epoch": 0.22580645161290322,
+      "grad_norm": 0.0,
+      "learning_rate": 9.874326750448835e-06,
+      "loss": 2.4969,
+      "step": 98
+    },
+    {
+      "epoch": 0.22811059907834103,
+      "grad_norm": 0.0,
+      "learning_rate": 9.872897676210041e-06,
+      "loss": 2.4532,
+      "step": 99
+    },
+    {
+      "epoch": 0.2304147465437788,
+      "grad_norm": 0.0,
+      "learning_rate": 9.871465295629821e-06,
+      "loss": 2.7851,
+      "step": 100
+    },
+    {
+      "epoch": 0.23271889400921658,
+      "grad_norm": 0.0,
+      "learning_rate": 9.870029597220437e-06,
+      "loss": 2.4056,
+      "step": 101
+    },
+    {
+      "epoch": 0.2350230414746544,
+      "grad_norm": 0.0,
+      "learning_rate": 9.868590569440867e-06,
+      "loss": 2.4184,
+      "step": 102
+    },
+    {
+      "epoch": 0.23732718894009217,
+      "grad_norm": 0.0,
+      "learning_rate": 9.867148200696505e-06,
+      "loss": 2.4005,
+      "step": 103
+    },
+    {
+      "epoch": 0.23963133640552994,
+      "grad_norm": 0.0,
+      "learning_rate": 9.865702479338843e-06,
+      "loss": 2.3063,
+      "step": 104
+    },
+    {
+      "epoch": 0.24193548387096775,
+      "grad_norm": 0.0,
+      "learning_rate": 9.86425339366516e-06,
+      "loss": 2.309,
+      "step": 105
+    },
+    {
+      "epoch": 0.24423963133640553,
+      "grad_norm": 0.0,
+      "learning_rate": 9.8628009319182e-06,
+      "loss": 2.2872,
+      "step": 106
+    },
+    {
+      "epoch": 0.2465437788018433,
+      "grad_norm": 0.0,
+      "learning_rate": 9.861345082285862e-06,
+      "loss": 2.5453,
+      "step": 107
+    },
+    {
+      "epoch": 0.2488479262672811,
+      "grad_norm": 0.0,
+      "learning_rate": 9.859885832900883e-06,
+      "loss": 2.4237,
+      "step": 108
+    },
+    {
+      "epoch": 0.2511520737327189,
+      "grad_norm": 0.0,
+      "learning_rate": 9.8584231718405e-06,
+      "loss": 2.7325,
+      "step": 109
+    },
+    {
+      "epoch": 0.2534562211981567,
+      "grad_norm": 0.0,
+      "learning_rate": 9.856957087126139e-06,
+      "loss": 2.8046,
+      "step": 110
+    },
+    {
+      "epoch": 0.2557603686635945,
+      "grad_norm": 0.0,
+      "learning_rate": 9.855487566723084e-06,
+      "loss": 2.3187,
+      "step": 111
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 0.0,
+      "learning_rate": 9.854014598540148e-06,
+      "loss": 2.1758,
+      "step": 112
+    },
+    {
+      "epoch": 0.26036866359447003,
+      "grad_norm": 0.0,
+      "learning_rate": 9.852538170429336e-06,
+      "loss": 2.3946,
+      "step": 113
+    },
+    {
+      "epoch": 0.2626728110599078,
+      "grad_norm": 0.0,
+      "learning_rate": 9.851058270185526e-06,
+      "loss": 2.5586,
+      "step": 114
+    },
+    {
+      "epoch": 0.26497695852534564,
+      "grad_norm": 0.0,
+      "learning_rate": 9.849574885546111e-06,
+      "loss": 2.4507,
+      "step": 115
+    },
+    {
+      "epoch": 0.2672811059907834,
+      "grad_norm": 0.0,
+      "learning_rate": 9.848088004190675e-06,
+      "loss": 2.2848,
+      "step": 116
+    },
+    {
+      "epoch": 0.2695852534562212,
+      "grad_norm": 0.0,
+      "learning_rate": 9.846597613740658e-06,
+      "loss": 2.7136,
+      "step": 117
+    },
+    {
+      "epoch": 0.271889400921659,
+      "grad_norm": 0.0,
+      "learning_rate": 9.845103701758992e-06,
+      "loss": 2.2407,
+      "step": 118
+    },
+    {
+      "epoch": 0.27419354838709675,
+      "grad_norm": 0.0,
+      "learning_rate": 9.843606255749772e-06,
+      "loss": 1.8917,
+      "step": 119
+    },
+    {
+      "epoch": 0.2764976958525346,
+      "grad_norm": 0.0,
+      "learning_rate": 9.842105263157896e-06,
+      "loss": 2.49,
+      "step": 120
+    },
+    {
+      "epoch": 0.27880184331797236,
+      "grad_norm": 0.0,
+      "learning_rate": 9.840600711368727e-06,
+      "loss": 2.2081,
+      "step": 121
+    },
+    {
+      "epoch": 0.28110599078341014,
+      "grad_norm": 0.0,
+      "learning_rate": 9.839092587707729e-06,
+      "loss": 2.5349,
+      "step": 122
+    },
+    {
+      "epoch": 0.2834101382488479,
+      "grad_norm": 0.0,
+      "learning_rate": 9.837580879440117e-06,
+      "loss": 2.2057,
+      "step": 123
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.0,
+      "learning_rate": 9.836065573770494e-06,
+      "loss": 2.6525,
+      "step": 124
+    },
+    {
+      "epoch": 0.2880184331797235,
+      "grad_norm": 0.0,
+      "learning_rate": 9.834546657842488e-06,
+      "loss": 2.466,
+      "step": 125
+    },
+    {
+      "epoch": 0.2903225806451613,
+      "grad_norm": 0.0,
+      "learning_rate": 9.833024118738406e-06,
+      "loss": 2.3602,
+      "step": 126
+    },
+    {
+      "epoch": 0.2926267281105991,
+      "grad_norm": 0.0,
+      "learning_rate": 9.831497943478838e-06,
+      "loss": 1.9779,
+      "step": 127
+    },
+    {
+      "epoch": 0.29493087557603687,
+      "grad_norm": 0.0,
+      "learning_rate": 9.829968119022318e-06,
+      "loss": 2.0746,
+      "step": 128
+    },
+    {
+      "epoch": 0.29723502304147464,
+      "grad_norm": 0.0,
+      "learning_rate": 9.82843463226493e-06,
+      "loss": 2.5047,
+      "step": 129
+    },
+    {
+      "epoch": 0.2995391705069124,
+      "grad_norm": 0.0,
+      "learning_rate": 9.826897470039947e-06,
+      "loss": 2.3455,
+      "step": 130
+    },
+    {
+      "epoch": 0.30184331797235026,
+      "grad_norm": 0.0,
+      "learning_rate": 9.825356619117452e-06,
+      "loss": 2.7969,
+      "step": 131
+    },
+    {
+      "epoch": 0.30414746543778803,
+      "grad_norm": 0.0,
+      "learning_rate": 9.823812066203951e-06,
+      "loss": 2.1092,
+      "step": 132
+    },
+    {
+      "epoch": 0.3064516129032258,
+      "grad_norm": 0.0,
+      "learning_rate": 9.822263797942004e-06,
+      "loss": 2.454,
+      "step": 133
+    },
+    {
+      "epoch": 0.3087557603686636,
+      "grad_norm": 0.0,
+      "learning_rate": 9.82071180090982e-06,
+      "loss": 2.2478,
+      "step": 134
+    },
+    {
+      "epoch": 0.31105990783410137,
+      "grad_norm": 0.0,
+      "learning_rate": 9.819156061620899e-06,
+      "loss": 2.1831,
+      "step": 135
+    },
+    {
+      "epoch": 0.31336405529953915,
+      "grad_norm": 0.0,
+      "learning_rate": 9.817596566523606e-06,
+      "loss": 2.199,
+      "step": 136
+    },
+    {
+      "epoch": 0.315668202764977,
+      "grad_norm": 0.0,
+      "learning_rate": 9.816033302000808e-06,
+      "loss": 2.3337,
+      "step": 137
+    },
+    {
+      "epoch": 0.31797235023041476,
+      "grad_norm": 0.0,
+      "learning_rate": 9.814466254369455e-06,
+      "loss": 2.3119,
+      "step": 138
+    },
+    {
+      "epoch": 0.32027649769585254,
+      "grad_norm": 0.0,
+      "learning_rate": 9.8128954098802e-06,
+      "loss": 2.319,
+      "step": 139
+    },
+    {
+      "epoch": 0.3225806451612903,
+      "grad_norm": 0.0,
+      "learning_rate": 9.811320754716981e-06,
+      "loss": 2.37,
+      "step": 140
+    },
+    {
+      "epoch": 0.3248847926267281,
+      "grad_norm": 0.0,
+      "learning_rate": 9.809742274996628e-06,
+      "loss": 2.3041,
+      "step": 141
+    },
+    {
+      "epoch": 0.3271889400921659,
+      "grad_norm": 0.0,
+      "learning_rate": 9.808159956768442e-06,
+      "loss": 2.7213,
+      "step": 142
+    },
+    {
+      "epoch": 0.3294930875576037,
+      "grad_norm": 0.0,
+      "learning_rate": 9.806573786013798e-06,
+      "loss": 2.6252,
+      "step": 143
+    },
+    {
+      "epoch": 0.3317972350230415,
+      "grad_norm": 0.0,
+      "learning_rate": 9.80498374864572e-06,
+      "loss": 1.8312,
+      "step": 144
+    },
+    {
+      "epoch": 0.33410138248847926,
+      "grad_norm": 0.0,
+      "learning_rate": 9.803389830508476e-06,
+      "loss": 2.4053,
+      "step": 145
+    },
+    {
+      "epoch": 0.33640552995391704,
+      "grad_norm": 0.0,
+      "learning_rate": 9.80179201737714e-06,
+      "loss": 2.3841,
+      "step": 146
+    },
+    {
+      "epoch": 0.3387096774193548,
+      "grad_norm": 0.0,
+      "learning_rate": 9.800190294957183e-06,
+      "loss": 2.4041,
+      "step": 147
+    },
+    {
+      "epoch": 0.34101382488479265,
+      "grad_norm": 0.0,
+      "learning_rate": 9.79858464888405e-06,
+      "loss": 2.0689,
+      "step": 148
+    },
+    {
+      "epoch": 0.3433179723502304,
+      "grad_norm": 0.0,
+      "learning_rate": 9.796975064722715e-06,
+      "loss": 2.0356,
+      "step": 149
+    },
+    {
+      "epoch": 0.3456221198156682,
+      "grad_norm": 0.0,
+      "learning_rate": 9.795361527967258e-06,
+      "loss": 2.0666,
+      "step": 150
+    },
+    {
+      "epoch": 0.347926267281106,
+      "grad_norm": 0.0,
+      "learning_rate": 9.793744024040433e-06,
+      "loss": 2.5768,
+      "step": 151
+    },
+    {
+      "epoch": 0.35023041474654376,
+      "grad_norm": 0.0,
+      "learning_rate": 9.792122538293217e-06,
+      "loss": 2.1563,
+      "step": 152
+    },
+    {
+      "epoch": 0.35253456221198154,
+      "grad_norm": 0.0,
+      "learning_rate": 9.790497056004383e-06,
+      "loss": 2.3727,
+      "step": 153
+    },
+    {
+      "epoch": 0.3548387096774194,
+      "grad_norm": 0.0,
+      "learning_rate": 9.78886756238004e-06,
+      "loss": 1.9836,
+      "step": 154
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 0.0,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 2.0014,
+      "step": 155
+    },
+    {
+      "epoch": 0.35944700460829493,
+      "grad_norm": 0.0,
+      "learning_rate": 9.785596481583288e-06,
+      "loss": 2.128,
+      "step": 156
+    },
+    {
+      "epoch": 0.3617511520737327,
+      "grad_norm": 0.0,
+      "learning_rate": 9.78395486445576e-06,
+      "loss": 1.9376,
+      "step": 157
+    },
+    {
+      "epoch": 0.3640552995391705,
+      "grad_norm": 0.0,
+      "learning_rate": 9.782309176081566e-06,
+      "loss": 2.6088,
+      "step": 158
+    },
+    {
+      "epoch": 0.3663594470046083,
+      "grad_norm": 0.0,
+      "learning_rate": 9.780659401296732e-06,
+      "loss": 2.476,
+      "step": 159
+    },
+    {
+      "epoch": 0.3686635944700461,
+      "grad_norm": 0.0,
+      "learning_rate": 9.779005524861878e-06,
+      "loss": 2.2163,
+      "step": 160
+    },
+    {
+      "epoch": 0.3709677419354839,
+      "grad_norm": 0.0,
+      "learning_rate": 9.777347531461764e-06,
+      "loss": 2.6048,
+      "step": 161
+    },
+    {
+      "epoch": 0.37327188940092165,
+      "grad_norm": 0.0,
+      "learning_rate": 9.775685405704793e-06,
+      "loss": 2.2061,
+      "step": 162
+    },
+    {
+      "epoch": 0.37557603686635943,
+      "grad_norm": 0.0,
+      "learning_rate": 9.774019132122557e-06,
+      "loss": 2.4101,
+      "step": 163
+    },
+    {
+      "epoch": 0.3778801843317972,
+      "grad_norm": 0.0,
+      "learning_rate": 9.77234869516935e-06,
+      "loss": 2.5347,
+      "step": 164
+    },
+    {
+      "epoch": 0.38018433179723504,
+      "grad_norm": 0.0,
+      "learning_rate": 9.770674079221682e-06,
+      "loss": 2.1867,
+      "step": 165
+    },
+    {
+      "epoch": 0.3824884792626728,
+      "grad_norm": 0.0,
+      "learning_rate": 9.768995268577792e-06,
+      "loss": 2.3605,
+      "step": 166
+    },
+    {
+      "epoch": 0.3847926267281106,
+      "grad_norm": 0.0,
+      "learning_rate": 9.767312247457155e-06,
+      "loss": 2.1918,
+      "step": 167
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 0.0,
+      "learning_rate": 9.765625e-06,
+      "loss": 2.4631,
+      "step": 168
+    },
+    {
+      "epoch": 0.38940092165898615,
+      "grad_norm": 0.0,
+      "learning_rate": 9.763933510266798e-06,
+      "loss": 2.077,
+      "step": 169
+    },
+    {
+      "epoch": 0.391705069124424,
+      "grad_norm": 0.0,
+      "learning_rate": 9.762237762237764e-06,
+      "loss": 2.274,
+      "step": 170
+    },
+    {
+      "epoch": 0.39400921658986177,
+      "grad_norm": 0.0,
+      "learning_rate": 9.760537739812353e-06,
+      "loss": 2.5203,
+      "step": 171
+    },
+    {
+      "epoch": 0.39631336405529954,
+      "grad_norm": 0.0,
+      "learning_rate": 9.75883342680875e-06,
+      "loss": 2.2468,
+      "step": 172
+    },
+    {
+      "epoch": 0.3986175115207373,
+      "grad_norm": 0.0,
+      "learning_rate": 9.757124806963358e-06,
+      "loss": 2.2579,
+      "step": 173
+    },
+    {
+      "epoch": 0.4009216589861751,
+      "grad_norm": 0.0,
+      "learning_rate": 9.755411863930279e-06,
+      "loss": 2.3655,
+      "step": 174
+    },
+    {
+      "epoch": 0.4032258064516129,
+      "grad_norm": 0.0,
+      "learning_rate": 9.75369458128079e-06,
+      "loss": 2.353,
+      "step": 175
+    },
+    {
+      "epoch": 0.4055299539170507,
+      "grad_norm": 0.0,
+      "learning_rate": 9.75197294250282e-06,
+      "loss": 2.2856,
+      "step": 176
+    },
+    {
+      "epoch": 0.4078341013824885,
+      "grad_norm": 0.0,
+      "learning_rate": 9.750246931000424e-06,
+      "loss": 2.2884,
+      "step": 177
+    },
+    {
+      "epoch": 0.41013824884792627,
+      "grad_norm": 0.0,
+      "learning_rate": 9.748516530093248e-06,
+      "loss": 2.1927,
+      "step": 178
+    },
+    {
+      "epoch": 0.41244239631336405,
+      "grad_norm": 0.0,
+      "learning_rate": 9.746781723015986e-06,
+      "loss": 2.4978,
+      "step": 179
+    },
+    {
+      "epoch": 0.4147465437788018,
+      "grad_norm": 0.0,
+      "learning_rate": 9.745042492917848e-06,
+      "loss": 2.1364,
+      "step": 180
+    },
+    {
+      "epoch": 0.41705069124423966,
+      "grad_norm": 0.0,
+      "learning_rate": 9.743298822862007e-06,
+      "loss": 2.4476,
+      "step": 181
+    },
+    {
+      "epoch": 0.41935483870967744,
+      "grad_norm": 0.0,
+      "learning_rate": 9.74155069582505e-06,
+      "loss": 1.8542,
+      "step": 182
+    },
+    {
+      "epoch": 0.4216589861751152,
+      "grad_norm": 0.0,
+      "learning_rate": 9.739798094696433e-06,
+      "loss": 2.2825,
+      "step": 183
+    },
+    {
+      "epoch": 0.423963133640553,
+      "grad_norm": 0.0,
+      "learning_rate": 9.738041002277906e-06,
+      "loss": 2.504,
+      "step": 184
+    },
+    {
+      "epoch": 0.42626728110599077,
+      "grad_norm": 0.0,
+      "learning_rate": 9.736279401282966e-06,
+      "loss": 2.1668,
+      "step": 185
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.0,
+      "learning_rate": 9.734513274336284e-06,
+      "loss": 2.2124,
+      "step": 186
+    },
+    {
+      "epoch": 0.4308755760368664,
+      "grad_norm": 0.0,
+      "learning_rate": 9.732742603973132e-06,
+      "loss": 2.0533,
+      "step": 187
+    },
+    {
+      "epoch": 0.43317972350230416,
+      "grad_norm": 0.0,
+      "learning_rate": 9.73096737263881e-06,
+      "loss": 2.2371,
+      "step": 188
+    },
+    {
+      "epoch": 0.43548387096774194,
+      "grad_norm": 0.0,
+      "learning_rate": 9.729187562688066e-06,
+      "loss": 2.577,
+      "step": 189
+    },
+    {
+      "epoch": 0.4377880184331797,
+      "grad_norm": 0.0,
+      "learning_rate": 9.727403156384506e-06,
+      "loss": 2.0062,
+      "step": 190
+    },
+    {
+      "epoch": 0.4400921658986175,
+      "grad_norm": 0.0,
+      "learning_rate": 9.725614135900015e-06,
+      "loss": 2.5623,
+      "step": 191
+    },
+    {
+      "epoch": 0.4423963133640553,
+      "grad_norm": 0.0,
+      "learning_rate": 9.723820483314155e-06,
+      "loss": 2.3879,
+      "step": 192
+    },
+    {
+      "epoch": 0.4447004608294931,
+      "grad_norm": 0.0,
+      "learning_rate": 9.722022180613568e-06,
+      "loss": 2.3884,
+      "step": 193
+    },
+    {
+      "epoch": 0.4470046082949309,
+      "grad_norm": 0.0,
+      "learning_rate": 9.720219209691377e-06,
+      "loss": 2.1677,
+      "step": 194
+    },
+    {
+      "epoch": 0.44930875576036866,
+      "grad_norm": 0.0,
+      "learning_rate": 9.71841155234657e-06,
+      "loss": 2.1849,
+      "step": 195
+    },
+    {
+      "epoch": 0.45161290322580644,
+      "grad_norm": 0.0,
+      "learning_rate": 9.7165991902834e-06,
+      "loss": 2.4744,
+      "step": 196
+    },
+    {
+      "epoch": 0.4539170506912442,
+      "grad_norm": 0.0,
+      "learning_rate": 9.714782105110758e-06,
+      "loss": 2.1845,
+      "step": 197
+    },
+    {
+      "epoch": 0.45622119815668205,
+      "grad_norm": 0.0,
+      "learning_rate": 9.71296027834155e-06,
+      "loss": 2.1075,
+      "step": 198
+    },
+    {
+      "epoch": 0.45852534562211983,
+      "grad_norm": 0.0,
+      "learning_rate": 9.711133691392076e-06,
+      "loss": 2.2594,
+      "step": 199
+    },
+    {
+      "epoch": 0.4608294930875576,
+      "grad_norm": 0.0,
+      "learning_rate": 9.709302325581395e-06,
+      "loss": 2.0445,
+      "step": 200
+    },
+    {
+      "epoch": 0.4631336405529954,
+      "grad_norm": 0.0,
+      "learning_rate": 9.707466162130695e-06,
+      "loss": 1.8306,
+      "step": 201
+    },
+    {
+      "epoch": 0.46543778801843316,
+      "grad_norm": 0.0,
+      "learning_rate": 9.705625182162635e-06,
+      "loss": 2.0827,
+      "step": 202
+    },
+    {
+      "epoch": 0.46774193548387094,
+      "grad_norm": 0.0,
+      "learning_rate": 9.703779366700715e-06,
+      "loss": 2.4998,
+      "step": 203
+    },
+    {
+      "epoch": 0.4700460829493088,
+      "grad_norm": 0.0,
+      "learning_rate": 9.701928696668615e-06,
+      "loss": 1.9692,
+      "step": 204
+    },
+    {
+      "epoch": 0.47235023041474655,
+      "grad_norm": 0.0,
+      "learning_rate": 9.70007315288954e-06,
+      "loss": 2.2138,
+      "step": 205
+    },
+    {
+      "epoch": 0.47465437788018433,
+      "grad_norm": 0.0,
+      "learning_rate": 9.698212716085556e-06,
+      "loss": 2.3699,
+      "step": 206
+    },
+    {
+      "epoch": 0.4769585253456221,
+      "grad_norm": 0.0,
+      "learning_rate": 9.696347366876926e-06,
+      "loss": 2.2658,
+      "step": 207
+    },
+    {
+      "epoch": 0.4792626728110599,
+      "grad_norm": 0.0,
+      "learning_rate": 9.694477085781434e-06,
+      "loss": 2.3592,
+      "step": 208
+    },
+    {
+      "epoch": 0.4815668202764977,
+      "grad_norm": 0.0,
+      "learning_rate": 9.692601853213708e-06,
+      "loss": 2.1424,
+      "step": 209
+    },
+    {
+      "epoch": 0.4838709677419355,
+      "grad_norm": 0.0,
+      "learning_rate": 9.690721649484536e-06,
+      "loss": 2.1127,
+      "step": 210
+    },
+    {
+      "epoch": 0.4861751152073733,
+      "grad_norm": 0.0,
+      "learning_rate": 9.688836454800178e-06,
+      "loss": 2.39,
+      "step": 211
+    },
+    {
+      "epoch": 0.48847926267281105,
+      "grad_norm": 0.0,
+      "learning_rate": 9.686946249261665e-06,
+      "loss": 2.2204,
+      "step": 212
+    },
+    {
+      "epoch": 0.49078341013824883,
+      "grad_norm": 0.0,
+      "learning_rate": 9.685051012864114e-06,
+      "loss": 2.2297,
+      "step": 213
+    },
+    {
+      "epoch": 0.4930875576036866,
+      "grad_norm": 0.0,
+      "learning_rate": 9.683150725496004e-06,
+      "loss": 2.2213,
+      "step": 214
+    },
+    {
+      "epoch": 0.49539170506912444,
+      "grad_norm": 0.0,
+      "learning_rate": 9.681245366938474e-06,
+      "loss": 2.3086,
+      "step": 215
+    },
+    {
+      "epoch": 0.4976958525345622,
+      "grad_norm": 0.0,
+      "learning_rate": 9.679334916864607e-06,
+      "loss": 2.1777,
+      "step": 216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 2.2379,
+      "step": 217
+    },
+    {
+      "epoch": 0.5023041474654378,
+      "grad_norm": 0.0,
+      "learning_rate": 9.675498660315571e-06,
+      "loss": 2.3132,
+      "step": 218
+    },
+    {
+      "epoch": 0.5046082949308756,
+      "grad_norm": 0.0,
+      "learning_rate": 9.673572812639739e-06,
+      "loss": 2.0896,
+      "step": 219
+    },
+    {
+      "epoch": 0.5069124423963134,
+      "grad_norm": 0.0,
+      "learning_rate": 9.671641791044779e-06,
+      "loss": 2.1555,
+      "step": 220
+    },
+    {
+      "epoch": 0.5092165898617511,
+      "grad_norm": 0.0,
+      "learning_rate": 9.669705574652519e-06,
+      "loss": 2.2675,
+      "step": 221
+    },
+    {
+      "epoch": 0.511520737327189,
+      "grad_norm": 0.0,
+      "learning_rate": 9.667764142472315e-06,
+      "loss": 2.0952,
+      "step": 222
+    },
+    {
+      "epoch": 0.5138248847926268,
+      "grad_norm": 0.0,
+      "learning_rate": 9.665817473400271e-06,
+      "loss": 2.1719,
+      "step": 223
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 0.0,
+      "learning_rate": 9.663865546218488e-06,
+      "loss": 2.269,
+      "step": 224
+    },
+    {
+      "epoch": 0.5184331797235023,
+      "grad_norm": 0.0,
+      "learning_rate": 9.66190833959429e-06,
+      "loss": 2.3643,
+      "step": 225
+    },
+    {
+      "epoch": 0.5207373271889401,
+      "grad_norm": 0.0,
+      "learning_rate": 9.659945832079446e-06,
+      "loss": 2.2851,
+      "step": 226
+    },
+    {
+      "epoch": 0.5230414746543779,
+      "grad_norm": 0.0,
+      "learning_rate": 9.657978002109388e-06,
+      "loss": 2.2314,
+      "step": 227
+    },
+    {
+      "epoch": 0.5253456221198156,
+      "grad_norm": 0.0,
+      "learning_rate": 9.656004828002415e-06,
+      "loss": 2.306,
+      "step": 228
+    },
+    {
+      "epoch": 0.5276497695852534,
+      "grad_norm": 0.0,
+      "learning_rate": 9.654026287958907e-06,
+      "loss": 2.72,
+      "step": 229
+    },
+    {
+      "epoch": 0.5299539170506913,
+      "grad_norm": 0.0,
+      "learning_rate": 9.652042360060516e-06,
+      "loss": 2.2472,
+      "step": 230
+    },
+    {
+      "epoch": 0.532258064516129,
+      "grad_norm": 0.0,
+      "learning_rate": 9.650053022269354e-06,
+      "loss": 2.14,
+      "step": 231
+    },
+    {
+      "epoch": 0.5345622119815668,
+      "grad_norm": 0.0,
+      "learning_rate": 9.648058252427186e-06,
+      "loss": 2.1352,
+      "step": 232
+    },
+    {
+      "epoch": 0.5368663594470046,
+      "grad_norm": 0.0,
+      "learning_rate": 9.646058028254597e-06,
+      "loss": 2.1396,
+      "step": 233
+    },
+    {
+      "epoch": 0.5391705069124424,
+      "grad_norm": 0.0,
+      "learning_rate": 9.644052327350167e-06,
+      "loss": 2.0091,
+      "step": 234
+    },
+    {
+      "epoch": 0.5414746543778802,
+      "grad_norm": 0.0,
+      "learning_rate": 9.642041127189643e-06,
+      "loss": 2.3559,
+      "step": 235
+    },
+    {
+      "epoch": 0.543778801843318,
+      "grad_norm": 0.0,
+      "learning_rate": 9.640024405125077e-06,
+      "loss": 2.2764,
+      "step": 236
+    },
+    {
+      "epoch": 0.5460829493087558,
+      "grad_norm": 0.0,
+      "learning_rate": 9.638002138383995e-06,
+      "loss": 2.4431,
+      "step": 237
+    },
+    {
+      "epoch": 0.5483870967741935,
+      "grad_norm": 0.0,
+      "learning_rate": 9.635974304068522e-06,
+      "loss": 2.2692,
+      "step": 238
+    },
+    {
+      "epoch": 0.5506912442396313,
+      "grad_norm": 0.0,
+      "learning_rate": 9.633940879154541e-06,
+      "loss": 2.2558,
+      "step": 239
+    },
+    {
+      "epoch": 0.5529953917050692,
+      "grad_norm": 0.0,
+      "learning_rate": 9.631901840490798e-06,
+      "loss": 1.8841,
+      "step": 240
+    },
+    {
+      "epoch": 0.5552995391705069,
+      "grad_norm": 0.0,
+      "learning_rate": 9.629857164798035e-06,
+      "loss": 2.243,
+      "step": 241
+    },
+    {
+      "epoch": 0.5576036866359447,
+      "grad_norm": 0.0,
+      "learning_rate": 9.627806828668103e-06,
+      "loss": 2.7055,
+      "step": 242
+    },
+    {
+      "epoch": 0.5599078341013825,
+      "grad_norm": 0.0,
+      "learning_rate": 9.625750808563067e-06,
+      "loss": 2.0709,
+      "step": 243
+    },
+    {
+      "epoch": 0.5622119815668203,
+      "grad_norm": 0.0,
+      "learning_rate": 9.623689080814313e-06,
+      "loss": 2.2729,
+      "step": 244
+    },
+    {
+      "epoch": 0.5645161290322581,
+      "grad_norm": 0.0,
+      "learning_rate": 9.621621621621622e-06,
+      "loss": 2.1815,
+      "step": 245
+    },
+    {
+      "epoch": 0.5668202764976958,
+      "grad_norm": 0.0,
+      "learning_rate": 9.619548407052275e-06,
+      "loss": 2.3074,
+      "step": 246
+    },
+    {
+      "epoch": 0.5691244239631337,
+      "grad_norm": 0.0,
+      "learning_rate": 9.617469413040112e-06,
+      "loss": 2.2204,
+      "step": 247
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.0,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 2.2254,
+      "step": 248
+    },
+    {
+      "epoch": 0.5737327188940092,
+      "grad_norm": 0.0,
+      "learning_rate": 9.613293989749963e-06,
+      "loss": 2.2028,
+      "step": 249
+    },
+    {
+      "epoch": 0.576036866359447,
+      "grad_norm": 0.0,
+      "learning_rate": 9.611197511664076e-06,
+      "loss": 1.9041,
+      "step": 250
+    },
+    {
+      "epoch": 0.5783410138248848,
+      "grad_norm": 0.0,
+      "learning_rate": 9.609095156517678e-06,
+      "loss": 2.1567,
+      "step": 251
+    },
+    {
+      "epoch": 0.5806451612903226,
+      "grad_norm": 0.0,
+      "learning_rate": 9.60698689956332e-06,
+      "loss": 2.2442,
+      "step": 252
+    },
+    {
+      "epoch": 0.5829493087557603,
+      "grad_norm": 0.0,
+      "learning_rate": 9.604872715914416e-06,
+      "loss": 2.244,
+      "step": 253
+    },
+    {
+      "epoch": 0.5852534562211982,
+      "grad_norm": 0.0,
+      "learning_rate": 9.602752580544262e-06,
+      "loss": 2.3902,
+      "step": 254
+    },
+    {
+      "epoch": 0.5875576036866359,
+      "grad_norm": 0.0,
+      "learning_rate": 9.600626468285045e-06,
+      "loss": 2.3403,
+      "step": 255
+    },
+    {
+      "epoch": 0.5898617511520737,
+      "grad_norm": 0.0,
+      "learning_rate": 9.598494353826851e-06,
+      "loss": 2.3575,
+      "step": 256
+    },
+    {
+      "epoch": 0.5921658986175116,
+      "grad_norm": 0.0,
+      "learning_rate": 9.596356211716666e-06,
+      "loss": 2.3505,
+      "step": 257
+    },
+    {
+      "epoch": 0.5944700460829493,
+      "grad_norm": 0.0,
+      "learning_rate": 9.594212016357346e-06,
+      "loss": 2.1773,
+      "step": 258
+    },
+    {
+      "epoch": 0.5967741935483871,
+      "grad_norm": 0.0,
+      "learning_rate": 9.592061742006616e-06,
+      "loss": 2.296,
+      "step": 259
+    },
+    {
+      "epoch": 0.5990783410138248,
+      "grad_norm": 0.0,
+      "learning_rate": 9.589905362776025e-06,
+      "loss": 1.9344,
+      "step": 260
+    },
+    {
+      "epoch": 0.6013824884792627,
+      "grad_norm": 0.0,
+      "learning_rate": 9.587742852629917e-06,
+      "loss": 2.5543,
+      "step": 261
+    },
+    {
+      "epoch": 0.6036866359447005,
+      "grad_norm": 0.0,
+      "learning_rate": 9.585574185384373e-06,
+      "loss": 2.7177,
+      "step": 262
+    },
+    {
+      "epoch": 0.6059907834101382,
+      "grad_norm": 0.0,
+      "learning_rate": 9.583399334706163e-06,
+      "loss": 2.3868,
+      "step": 263
+    },
+    {
+      "epoch": 0.6082949308755761,
+      "grad_norm": 0.0,
+      "learning_rate": 9.581218274111676e-06,
+      "loss": 2.3185,
+      "step": 264
+    },
+    {
+      "epoch": 0.6105990783410138,
+      "grad_norm": 0.0,
+      "learning_rate": 9.579030976965846e-06,
+      "loss": 1.9385,
+      "step": 265
+    },
+    {
+      "epoch": 0.6129032258064516,
+      "grad_norm": 0.0,
+      "learning_rate": 9.57683741648107e-06,
+      "loss": 2.3687,
+      "step": 266
+    },
+    {
+      "epoch": 0.6152073732718893,
+      "grad_norm": 0.0,
+      "learning_rate": 9.574637565716107e-06,
+      "loss": 2.2045,
+      "step": 267
+    },
+    {
+      "epoch": 0.6175115207373272,
+      "grad_norm": 0.0,
+      "learning_rate": 9.572431397574987e-06,
+      "loss": 2.4644,
+      "step": 268
+    },
+    {
+      "epoch": 0.619815668202765,
+      "grad_norm": 0.0,
+      "learning_rate": 9.57021888480588e-06,
+      "loss": 2.3897,
+      "step": 269
+    },
+    {
+      "epoch": 0.6221198156682027,
+      "grad_norm": 0.0,
+      "learning_rate": 9.568e-06,
+      "loss": 2.4176,
+      "step": 270
+    },
+    {
+      "epoch": 0.6244239631336406,
+      "grad_norm": 0.0,
+      "learning_rate": 9.56577471559045e-06,
+      "loss": 2.3342,
+      "step": 271
+    },
+    {
+      "epoch": 0.6267281105990783,
+      "grad_norm": 0.0,
+      "learning_rate": 9.563543003851093e-06,
+      "loss": 2.3815,
+      "step": 272
+    },
+    {
+      "epoch": 0.6290322580645161,
+      "grad_norm": 0.0,
+      "learning_rate": 9.561304836895388e-06,
+      "loss": 2.435,
+      "step": 273
+    },
+    {
+      "epoch": 0.631336405529954,
+      "grad_norm": 0.0,
+      "learning_rate": 9.55906018667525e-06,
+      "loss": 2.4491,
+      "step": 274
+    },
+    {
+      "epoch": 0.6336405529953917,
+      "grad_norm": 0.0,
+      "learning_rate": 9.556809024979856e-06,
+      "loss": 2.3206,
+      "step": 275
+    },
+    {
+      "epoch": 0.6359447004608295,
+      "grad_norm": 0.0,
+      "learning_rate": 9.554551323434475e-06,
+      "loss": 2.1149,
+      "step": 276
+    },
+    {
+      "epoch": 0.6382488479262672,
+      "grad_norm": 0.0,
+      "learning_rate": 9.552287053499273e-06,
+      "loss": 2.2707,
+      "step": 277
+    },
+    {
+      "epoch": 0.6405529953917051,
+      "grad_norm": 0.0,
+      "learning_rate": 9.550016186468114e-06,
+      "loss": 2.3246,
+      "step": 278
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 0.0,
+      "learning_rate": 9.547738693467338e-06,
+      "loss": 2.2557,
+      "step": 279
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 0.0,
+      "learning_rate": 9.545454545454547e-06,
+      "loss": 2.3648,
+      "step": 280
+    },
+    {
+      "epoch": 0.6474654377880185,
+      "grad_norm": 0.0,
+      "learning_rate": 9.543163713217365e-06,
+      "loss": 2.3793,
+      "step": 281
+    },
+    {
+      "epoch": 0.6497695852534562,
+      "grad_norm": 0.0,
+      "learning_rate": 9.540866167372192e-06,
+      "loss": 2.2765,
+      "step": 282
+    },
+    {
+      "epoch": 0.652073732718894,
+      "grad_norm": 0.0,
+      "learning_rate": 9.538561878362955e-06,
+      "loss": 2.3297,
+      "step": 283
+    },
+    {
+      "epoch": 0.6543778801843319,
+      "grad_norm": 0.0,
+      "learning_rate": 9.53625081645983e-06,
+      "loss": 2.2636,
+      "step": 284
+    },
+    {
+      "epoch": 0.6566820276497696,
+      "grad_norm": 0.0,
+      "learning_rate": 9.533932951757974e-06,
+      "loss": 2.271,
+      "step": 285
+    },
+    {
+      "epoch": 0.6589861751152074,
+      "grad_norm": 0.0,
+      "learning_rate": 9.53160825417622e-06,
+      "loss": 2.497,
+      "step": 286
+    },
+    {
+      "epoch": 0.6612903225806451,
+      "grad_norm": 0.0,
+      "learning_rate": 9.529276693455799e-06,
+      "loss": 1.994,
+      "step": 287
+    },
+    {
+      "epoch": 0.663594470046083,
+      "grad_norm": 0.0,
+      "learning_rate": 9.526938239159003e-06,
+      "loss": 1.9041,
+      "step": 288
+    },
+    {
+      "epoch": 0.6658986175115207,
+      "grad_norm": 0.0,
+      "learning_rate": 9.524592860667874e-06,
+      "loss": 2.3384,
+      "step": 289
+    },
+    {
+      "epoch": 0.6682027649769585,
+      "grad_norm": 0.0,
+      "learning_rate": 9.522240527182867e-06,
+      "loss": 2.0346,
+      "step": 290
+    },
+    {
+      "epoch": 0.6705069124423964,
+      "grad_norm": 0.0,
+      "learning_rate": 9.519881207721498e-06,
+      "loss": 2.1251,
+      "step": 291
+    },
+    {
+      "epoch": 0.6728110599078341,
+      "grad_norm": 0.0,
+      "learning_rate": 9.517514871116987e-06,
+      "loss": 2.3156,
+      "step": 292
+    },
+    {
+      "epoch": 0.6751152073732719,
+      "grad_norm": 0.0,
+      "learning_rate": 9.51514148601688e-06,
+      "loss": 2.3411,
+      "step": 293
+    },
+    {
+      "epoch": 0.6774193548387096,
+      "grad_norm": 0.0,
+      "learning_rate": 9.512761020881673e-06,
+      "loss": 2.5454,
+      "step": 294
+    },
+    {
+      "epoch": 0.6797235023041475,
+      "grad_norm": 0.0,
+      "learning_rate": 9.510373443983403e-06,
+      "loss": 1.864,
+      "step": 295
+    },
+    {
+      "epoch": 0.6820276497695853,
+      "grad_norm": 0.0,
+      "learning_rate": 9.507978723404256e-06,
+      "loss": 2.2479,
+      "step": 296
+    },
+    {
+      "epoch": 0.684331797235023,
+      "grad_norm": 0.0,
+      "learning_rate": 9.505576827035126e-06,
+      "loss": 2.2826,
+      "step": 297
+    },
+    {
+      "epoch": 0.6866359447004609,
+      "grad_norm": 0.0,
+      "learning_rate": 9.503167722574192e-06,
+      "loss": 1.8908,
+      "step": 298
+    },
+    {
+      "epoch": 0.6889400921658986,
+      "grad_norm": 0.0,
+      "learning_rate": 9.500751377525463e-06,
+      "loss": 2.0887,
+      "step": 299
+    },
+    {
+      "epoch": 0.6912442396313364,
+      "grad_norm": 0.0,
+      "learning_rate": 9.498327759197325e-06,
+      "loss": 1.9868,
+      "step": 300
+    },
+    {
+      "epoch": 0.6935483870967742,
+      "grad_norm": 0.0,
+      "learning_rate": 9.495896834701056e-06,
+      "loss": 2.03,
+      "step": 301
+    },
+    {
+      "epoch": 0.695852534562212,
+      "grad_norm": 0.0,
+      "learning_rate": 9.493458570949348e-06,
+      "loss": 2.413,
+      "step": 302
+    },
+    {
+      "epoch": 0.6981566820276498,
+      "grad_norm": 0.0,
+      "learning_rate": 9.491012934654798e-06,
+      "loss": 2.2828,
+      "step": 303
+    },
+    {
+      "epoch": 0.7004608294930875,
+      "grad_norm": 0.0,
+      "learning_rate": 9.488559892328399e-06,
+      "loss": 2.4673,
+      "step": 304
+    },
+    {
+      "epoch": 0.7027649769585254,
+      "grad_norm": 0.0,
+      "learning_rate": 9.486099410278012e-06,
+      "loss": 2.1999,
+      "step": 305
+    },
+    {
+      "epoch": 0.7050691244239631,
+      "grad_norm": 0.0,
+      "learning_rate": 9.483631454606818e-06,
+      "loss": 2.1067,
+      "step": 306
+    },
+    {
+      "epoch": 0.7073732718894009,
+      "grad_norm": 0.0,
+      "learning_rate": 9.481155991211765e-06,
+      "loss": 2.0962,
+      "step": 307
+    },
+    {
+      "epoch": 0.7096774193548387,
+      "grad_norm": 0.0,
+      "learning_rate": 9.47867298578199e-06,
+      "loss": 2.4821,
+      "step": 308
+    },
+    {
+      "epoch": 0.7119815668202765,
+      "grad_norm": 0.0,
+      "learning_rate": 9.476182403797255e-06,
+      "loss": 2.4688,
+      "step": 309
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.0,
+      "learning_rate": 9.473684210526317e-06,
+      "loss": 2.1085,
+      "step": 310
+    },
+    {
+      "epoch": 0.716589861751152,
+      "grad_norm": 0.0,
+      "learning_rate": 9.471178371025337e-06,
+      "loss": 2.435,
+      "step": 311
+    },
+    {
+      "epoch": 0.7188940092165899,
+      "grad_norm": 0.0,
+      "learning_rate": 9.46866485013624e-06,
+      "loss": 2.3131,
+      "step": 312
+    },
+    {
+      "epoch": 0.7211981566820277,
+      "grad_norm": 0.0,
+      "learning_rate": 9.466143612485077e-06,
+      "loss": 2.5537,
+      "step": 313
+    },
+    {
+      "epoch": 0.7235023041474654,
+      "grad_norm": 0.0,
+      "learning_rate": 9.463614622480356e-06,
+      "loss": 2.1758,
+      "step": 314
+    },
+    {
+      "epoch": 0.7258064516129032,
+      "grad_norm": 0.0,
+      "learning_rate": 9.461077844311378e-06,
+      "loss": 2.2752,
+      "step": 315
+    },
+    {
+      "epoch": 0.728110599078341,
+      "grad_norm": 0.0,
+      "learning_rate": 9.45853324194654e-06,
+      "loss": 2.0488,
+      "step": 316
+    },
+    {
+      "epoch": 0.7304147465437788,
+      "grad_norm": 0.0,
+      "learning_rate": 9.45598077913163e-06,
+      "loss": 2.3105,
+      "step": 317
+    },
+    {
+      "epoch": 0.7327188940092166,
+      "grad_norm": 0.0,
+      "learning_rate": 9.453420419388106e-06,
+      "loss": 1.936,
+      "step": 318
+    },
+    {
+      "epoch": 0.7350230414746544,
+      "grad_norm": 0.0,
+      "learning_rate": 9.450852126011364e-06,
+      "loss": 2.1634,
+      "step": 319
+    },
+    {
+      "epoch": 0.7373271889400922,
+      "grad_norm": 0.0,
+      "learning_rate": 9.448275862068967e-06,
+      "loss": 1.9953,
+      "step": 320
+    },
+    {
+      "epoch": 0.7396313364055299,
+      "grad_norm": 0.0,
+      "learning_rate": 9.445691590398895e-06,
+      "loss": 2.3994,
+      "step": 321
+    },
+    {
+      "epoch": 0.7419354838709677,
+      "grad_norm": 0.0,
+      "learning_rate": 9.44309927360775e-06,
+      "loss": 1.837,
+      "step": 322
+    },
+    {
+      "epoch": 0.7442396313364056,
+      "grad_norm": 0.0,
+      "learning_rate": 9.440498874068943e-06,
+      "loss": 2.0258,
+      "step": 323
+    },
+    {
+      "epoch": 0.7465437788018433,
+      "grad_norm": 0.0,
+      "learning_rate": 9.43789035392089e-06,
+      "loss": 2.5129,
+      "step": 324
+    },
+    {
+      "epoch": 0.7488479262672811,
+      "grad_norm": 0.0,
+      "learning_rate": 9.435273675065161e-06,
+      "loss": 2.0436,
+      "step": 325
+    },
+    {
+      "epoch": 0.7511520737327189,
+      "grad_norm": 0.0,
+      "learning_rate": 9.432648799164637e-06,
+      "loss": 2.4449,
+      "step": 326
+    },
+    {
+      "epoch": 0.7534562211981567,
+      "grad_norm": 0.0,
+      "learning_rate": 9.430015687641625e-06,
+      "loss": 2.0828,
+      "step": 327
+    },
+    {
+      "epoch": 0.7557603686635944,
+      "grad_norm": 0.0,
+      "learning_rate": 9.42737430167598e-06,
+      "loss": 2.3661,
+      "step": 328
+    },
+    {
+      "epoch": 0.7580645161290323,
+      "grad_norm": 0.0,
+      "learning_rate": 9.424724602203184e-06,
+      "loss": 1.9816,
+      "step": 329
+    },
+    {
+      "epoch": 0.7603686635944701,
+      "grad_norm": 0.0,
+      "learning_rate": 9.422066549912434e-06,
+      "loss": 2.3207,
+      "step": 330
+    },
+    {
+      "epoch": 0.7626728110599078,
+      "grad_norm": 0.0,
+      "learning_rate": 9.419400105244694e-06,
+      "loss": 2.1479,
+      "step": 331
+    },
+    {
+      "epoch": 0.7649769585253456,
+      "grad_norm": 0.0,
+      "learning_rate": 9.416725228390726e-06,
+      "loss": 2.211,
+      "step": 332
+    },
+    {
+      "epoch": 0.7672811059907834,
+      "grad_norm": 0.0,
+      "learning_rate": 9.41404187928911e-06,
+      "loss": 1.8804,
+      "step": 333
+    },
+    {
+      "epoch": 0.7695852534562212,
+      "grad_norm": 0.0,
+      "learning_rate": 9.411350017624251e-06,
+      "loss": 2.1775,
+      "step": 334
+    },
+    {
+      "epoch": 0.771889400921659,
+      "grad_norm": 0.0,
+      "learning_rate": 9.40864960282436e-06,
+      "loss": 2.2466,
+      "step": 335
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 0.0,
+      "learning_rate": 9.405940594059407e-06,
+      "loss": 2.0984,
+      "step": 336
+    },
+    {
+      "epoch": 0.7764976958525346,
+      "grad_norm": 0.0,
+      "learning_rate": 9.403222950239065e-06,
+      "loss": 2.5951,
+      "step": 337
+    },
+    {
+      "epoch": 0.7788018433179723,
+      "grad_norm": 0.0,
+      "learning_rate": 9.400496630010643e-06,
+      "loss": 2.1321,
+      "step": 338
+    },
+    {
+      "epoch": 0.7811059907834101,
+      "grad_norm": 0.0,
+      "learning_rate": 9.397761591756973e-06,
+      "loss": 1.8965,
+      "step": 339
+    },
+    {
+      "epoch": 0.783410138248848,
+      "grad_norm": 0.0,
+      "learning_rate": 9.395017793594307e-06,
+      "loss": 2.4059,
+      "step": 340
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 0.0,
+      "learning_rate": 9.392265193370167e-06,
+      "loss": 2.4522,
+      "step": 341
+    },
+    {
+      "epoch": 0.7880184331797235,
+      "grad_norm": 0.0,
+      "learning_rate": 9.389503748661194e-06,
+      "loss": 2.4031,
+      "step": 342
+    },
+    {
+      "epoch": 0.7903225806451613,
+      "grad_norm": 0.0,
+      "learning_rate": 9.386733416770964e-06,
+      "loss": 2.3547,
+      "step": 343
+    },
+    {
+      "epoch": 0.7926267281105991,
+      "grad_norm": 0.0,
+      "learning_rate": 9.383954154727794e-06,
+      "loss": 2.0233,
+      "step": 344
+    },
+    {
+      "epoch": 0.7949308755760369,
+      "grad_norm": 0.0,
+      "learning_rate": 9.381165919282512e-06,
+      "loss": 2.3673,
+      "step": 345
+    },
+    {
+      "epoch": 0.7972350230414746,
+      "grad_norm": 0.0,
+      "learning_rate": 9.378368666906218e-06,
+      "loss": 2.4832,
+      "step": 346
+    },
+    {
+      "epoch": 0.7995391705069125,
+      "grad_norm": 0.0,
+      "learning_rate": 9.375562353788017e-06,
+      "loss": 2.0684,
+      "step": 347
+    },
+    {
+      "epoch": 0.8018433179723502,
+      "grad_norm": 0.0,
+      "learning_rate": 9.372746935832732e-06,
+      "loss": 2.1924,
+      "step": 348
+    },
+    {
+      "epoch": 0.804147465437788,
+      "grad_norm": 0.0,
+      "learning_rate": 9.369922368658603e-06,
+      "loss": 2.3609,
+      "step": 349
+    },
+    {
+      "epoch": 0.8064516129032258,
+      "grad_norm": 0.0,
+      "learning_rate": 9.367088607594939e-06,
+      "loss": 2.2614,
+      "step": 350
+    },
+    {
+      "epoch": 0.8087557603686636,
+      "grad_norm": 0.0,
+      "learning_rate": 9.364245607679769e-06,
+      "loss": 2.5841,
+      "step": 351
+    },
+    {
+      "epoch": 0.8110599078341014,
+      "grad_norm": 0.0,
+      "learning_rate": 9.361393323657475e-06,
+      "loss": 2.038,
+      "step": 352
+    },
+    {
+      "epoch": 0.8133640552995391,
+      "grad_norm": 0.0,
+      "learning_rate": 9.358531709976377e-06,
+      "loss": 2.2917,
+      "step": 353
+    },
+    {
+      "epoch": 0.815668202764977,
+      "grad_norm": 0.0,
+      "learning_rate": 9.355660720786313e-06,
+      "loss": 2.0844,
+      "step": 354
+    },
+    {
+      "epoch": 0.8179723502304147,
+      "grad_norm": 0.0,
+      "learning_rate": 9.352780309936191e-06,
+      "loss": 1.7098,
+      "step": 355
+    },
+    {
+      "epoch": 0.8202764976958525,
+      "grad_norm": 0.0,
+      "learning_rate": 9.349890430971511e-06,
+      "loss": 2.4264,
+      "step": 356
+    },
+    {
+      "epoch": 0.8225806451612904,
+      "grad_norm": 0.0,
+      "learning_rate": 9.346991037131882e-06,
+      "loss": 2.4563,
+      "step": 357
+    },
+    {
+      "epoch": 0.8248847926267281,
+      "grad_norm": 0.0,
+      "learning_rate": 9.34408208134848e-06,
+      "loss": 1.9052,
+      "step": 358
+    },
+    {
+      "epoch": 0.8271889400921659,
+      "grad_norm": 0.0,
+      "learning_rate": 9.341163516241514e-06,
+      "loss": 2.1467,
+      "step": 359
+    },
+    {
+      "epoch": 0.8294930875576036,
+      "grad_norm": 0.0,
+      "learning_rate": 9.338235294117646e-06,
+      "loss": 2.2446,
+      "step": 360
+    },
+    {
+      "epoch": 0.8317972350230415,
+      "grad_norm": 0.0,
+      "learning_rate": 9.33529736696741e-06,
+      "loss": 2.1227,
+      "step": 361
+    },
+    {
+      "epoch": 0.8341013824884793,
+      "grad_norm": 0.0,
+      "learning_rate": 9.332349686462561e-06,
+      "loss": 2.2938,
+      "step": 362
+    },
+    {
+      "epoch": 0.836405529953917,
+      "grad_norm": 0.0,
+      "learning_rate": 9.329392203953447e-06,
+      "loss": 2.2272,
+      "step": 363
+    },
+    {
+      "epoch": 0.8387096774193549,
+      "grad_norm": 0.0,
+      "learning_rate": 9.326424870466322e-06,
+      "loss": 1.9074,
+      "step": 364
+    },
+    {
+      "epoch": 0.8410138248847926,
+      "grad_norm": 0.0,
+      "learning_rate": 9.32344763670065e-06,
+      "loss": 1.7826,
+      "step": 365
+    },
+    {
+      "epoch": 0.8433179723502304,
+      "grad_norm": 0.0,
+      "learning_rate": 9.320460453026365e-06,
+      "loss": 2.4799,
+      "step": 366
+    },
+    {
+      "epoch": 0.8456221198156681,
+      "grad_norm": 0.0,
+      "learning_rate": 9.317463269481124e-06,
+      "loss": 1.8732,
+      "step": 367
+    },
+    {
+      "epoch": 0.847926267281106,
+      "grad_norm": 0.0,
+      "learning_rate": 9.314456035767513e-06,
+      "loss": 2.4681,
+      "step": 368
+    },
+    {
+      "epoch": 0.8502304147465438,
+      "grad_norm": 0.0,
+      "learning_rate": 9.311438701250233e-06,
+      "loss": 1.7413,
+      "step": 369
+    },
+    {
+      "epoch": 0.8525345622119815,
+      "grad_norm": 0.0,
+      "learning_rate": 9.308411214953271e-06,
+      "loss": 2.2242,
+      "step": 370
+    },
+    {
+      "epoch": 0.8548387096774194,
+      "grad_norm": 0.0,
+      "learning_rate": 9.305373525557013e-06,
+      "loss": 2.6895,
+      "step": 371
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.0,
+      "learning_rate": 9.30232558139535e-06,
+      "loss": 1.8005,
+      "step": 372
+    },
+    {
+      "epoch": 0.8594470046082949,
+      "grad_norm": 0.0,
+      "learning_rate": 9.299267330452753e-06,
+      "loss": 1.9613,
+      "step": 373
+    },
+    {
+      "epoch": 0.8617511520737328,
+      "grad_norm": 0.0,
+      "learning_rate": 9.29619872036131e-06,
+      "loss": 2.441,
+      "step": 374
+    },
+    {
+      "epoch": 0.8640552995391705,
+      "grad_norm": 0.0,
+      "learning_rate": 9.29311969839774e-06,
+      "loss": 2.0829,
+      "step": 375
+    },
+    {
+      "epoch": 0.8663594470046083,
+      "grad_norm": 0.0,
+      "learning_rate": 9.290030211480364e-06,
+      "loss": 2.0612,
+      "step": 376
+    },
+    {
+      "epoch": 0.868663594470046,
+      "grad_norm": 0.0,
+      "learning_rate": 9.286930206166069e-06,
+      "loss": 2.1892,
+      "step": 377
+    },
+    {
+      "epoch": 0.8709677419354839,
+      "grad_norm": 0.0,
+      "learning_rate": 9.283819628647214e-06,
+      "loss": 2.562,
+      "step": 378
+    },
+    {
+      "epoch": 0.8732718894009217,
+      "grad_norm": 0.0,
+      "learning_rate": 9.280698424748529e-06,
+      "loss": 2.2855,
+      "step": 379
+    },
+    {
+      "epoch": 0.8755760368663594,
+      "grad_norm": 0.0,
+      "learning_rate": 9.277566539923955e-06,
+      "loss": 2.4445,
+      "step": 380
+    },
+    {
+      "epoch": 0.8778801843317973,
+      "grad_norm": 0.0,
+      "learning_rate": 9.274423919253477e-06,
+      "loss": 2.1253,
+      "step": 381
+    },
+    {
+      "epoch": 0.880184331797235,
+      "grad_norm": 0.0,
+      "learning_rate": 9.271270507439909e-06,
+      "loss": 2.4146,
+      "step": 382
+    },
+    {
+      "epoch": 0.8824884792626728,
+      "grad_norm": 0.0,
+      "learning_rate": 9.268106248805657e-06,
+      "loss": 2.4259,
+      "step": 383
+    },
+    {
+      "epoch": 0.8847926267281107,
+      "grad_norm": 0.0,
+      "learning_rate": 9.264931087289435e-06,
+      "loss": 2.227,
+      "step": 384
+    },
+    {
+      "epoch": 0.8870967741935484,
+      "grad_norm": 0.0,
+      "learning_rate": 9.261744966442955e-06,
+      "loss": 2.437,
+      "step": 385
+    },
+    {
+      "epoch": 0.8894009216589862,
+      "grad_norm": 0.0,
+      "learning_rate": 9.258547829427584e-06,
+      "loss": 2.3971,
+      "step": 386
+    },
+    {
+      "epoch": 0.8917050691244239,
+      "grad_norm": 0.0,
+      "learning_rate": 9.255339619010967e-06,
+      "loss": 2.164,
+      "step": 387
+    },
+    {
+      "epoch": 0.8940092165898618,
+      "grad_norm": 0.0,
+      "learning_rate": 9.252120277563609e-06,
+      "loss": 2.22,
+      "step": 388
+    },
+    {
+      "epoch": 0.8963133640552995,
+      "grad_norm": 0.0,
+      "learning_rate": 9.248889747055418e-06,
+      "loss": 2.1684,
+      "step": 389
+    },
+    {
+      "epoch": 0.8986175115207373,
+      "grad_norm": 0.0,
+      "learning_rate": 9.245647969052225e-06,
+      "loss": 2.0864,
+      "step": 390
+    },
+    {
+      "epoch": 0.9009216589861752,
+      "grad_norm": 0.0,
+      "learning_rate": 9.242394884712266e-06,
+      "loss": 2.3236,
+      "step": 391
+    },
+    {
+      "epoch": 0.9032258064516129,
+      "grad_norm": 0.0,
+      "learning_rate": 9.23913043478261e-06,
+      "loss": 2.205,
+      "step": 392
+    },
+    {
+      "epoch": 0.9055299539170507,
+      "grad_norm": 0.0,
+      "learning_rate": 9.235854559595567e-06,
+      "loss": 1.9739,
+      "step": 393
+    },
+    {
+      "epoch": 0.9078341013824884,
+      "grad_norm": 0.0,
+      "learning_rate": 9.232567199065057e-06,
+      "loss": 2.1698,
+      "step": 394
+    },
+    {
+      "epoch": 0.9101382488479263,
+      "grad_norm": 0.0,
+      "learning_rate": 9.229268292682927e-06,
+      "loss": 2.1516,
+      "step": 395
+    },
+    {
+      "epoch": 0.9124423963133641,
+      "grad_norm": 0.0,
+      "learning_rate": 9.225957779515246e-06,
+      "loss": 1.8965,
+      "step": 396
+    },
+    {
+      "epoch": 0.9147465437788018,
+      "grad_norm": 0.0,
+      "learning_rate": 9.222635598198551e-06,
+      "loss": 2.0203,
+      "step": 397
+    },
+    {
+      "epoch": 0.9170506912442397,
+      "grad_norm": 0.0,
+      "learning_rate": 9.219301686936056e-06,
+      "loss": 2.5318,
+      "step": 398
+    },
+    {
+      "epoch": 0.9193548387096774,
+      "grad_norm": 0.0,
+      "learning_rate": 9.21595598349381e-06,
+      "loss": 2.1496,
+      "step": 399
+    },
+    {
+      "epoch": 0.9216589861751152,
+      "grad_norm": 0.0,
+      "learning_rate": 9.21259842519685e-06,
+      "loss": 1.8457,
+      "step": 400
+    },
+    {
+      "epoch": 0.923963133640553,
+      "grad_norm": 0.0,
+      "learning_rate": 9.209228948925261e-06,
+      "loss": 2.0916,
+      "step": 401
+    },
+    {
+      "epoch": 0.9262672811059908,
+      "grad_norm": 0.0,
+      "learning_rate": 9.205847491110234e-06,
+      "loss": 2.4179,
+      "step": 402
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 0.0,
+      "learning_rate": 9.202453987730062e-06,
+      "loss": 1.8332,
+      "step": 403
+    },
+    {
+      "epoch": 0.9308755760368663,
+      "grad_norm": 0.0,
+      "learning_rate": 9.199048374306107e-06,
+      "loss": 2.5119,
+      "step": 404
+    },
+    {
+      "epoch": 0.9331797235023042,
+      "grad_norm": 0.0,
+      "learning_rate": 9.19563058589871e-06,
+      "loss": 1.9103,
+      "step": 405
+    },
+    {
+      "epoch": 0.9354838709677419,
+      "grad_norm": 0.0,
+      "learning_rate": 9.192200557103064e-06,
+      "loss": 2.5436,
+      "step": 406
+    },
+    {
+      "epoch": 0.9377880184331797,
+      "grad_norm": 0.0,
+      "learning_rate": 9.188758222045046e-06,
+      "loss": 2.1996,
+      "step": 407
+    },
+    {
+      "epoch": 0.9400921658986175,
+      "grad_norm": 0.0,
+      "learning_rate": 9.185303514376996e-06,
+      "loss": 2.1928,
+      "step": 408
+    },
+    {
+      "epoch": 0.9423963133640553,
+      "grad_norm": 0.0,
+      "learning_rate": 9.181836367273456e-06,
+      "loss": 2.39,
+      "step": 409
+    },
+    {
+      "epoch": 0.9447004608294931,
+      "grad_norm": 0.0,
+      "learning_rate": 9.178356713426856e-06,
+      "loss": 1.9289,
+      "step": 410
+    },
+    {
+      "epoch": 0.9470046082949308,
+      "grad_norm": 0.0,
+      "learning_rate": 9.174864485043165e-06,
+      "loss": 1.6711,
+      "step": 411
+    },
+    {
+      "epoch": 0.9493087557603687,
+      "grad_norm": 0.0,
+      "learning_rate": 9.17135961383749e-06,
+      "loss": 1.9521,
+      "step": 412
+    },
+    {
+      "epoch": 0.9516129032258065,
+      "grad_norm": 0.0,
+      "learning_rate": 9.16784203102962e-06,
+      "loss": 2.1426,
+      "step": 413
+    },
+    {
+      "epoch": 0.9539170506912442,
+      "grad_norm": 0.0,
+      "learning_rate": 9.164311667339525e-06,
+      "loss": 2.3366,
+      "step": 414
+    },
+    {
+      "epoch": 0.956221198156682,
+      "grad_norm": 0.0,
+      "learning_rate": 9.160768452982811e-06,
+      "loss": 2.5255,
+      "step": 415
+    },
+    {
+      "epoch": 0.9585253456221198,
+      "grad_norm": 0.0,
+      "learning_rate": 9.157212317666126e-06,
+      "loss": 2.3075,
+      "step": 416
+    },
+    {
+      "epoch": 0.9608294930875576,
+      "grad_norm": 0.0,
+      "learning_rate": 9.153643190582505e-06,
+      "loss": 2.2583,
+      "step": 417
+    },
+    {
+      "epoch": 0.9631336405529954,
+      "grad_norm": 0.0,
+      "learning_rate": 9.15006100040667e-06,
+      "loss": 2.0314,
+      "step": 418
+    },
+    {
+      "epoch": 0.9654377880184332,
+      "grad_norm": 0.0,
+      "learning_rate": 9.146465675290284e-06,
+      "loss": 2.3014,
+      "step": 419
+    },
+    {
+      "epoch": 0.967741935483871,
+      "grad_norm": 0.0,
+      "learning_rate": 9.142857142857142e-06,
+      "loss": 2.5585,
+      "step": 420
+    },
+    {
+      "epoch": 0.9700460829493087,
+      "grad_norm": 0.0,
+      "learning_rate": 9.139235330198323e-06,
+      "loss": 2.2314,
+      "step": 421
+    },
+    {
+      "epoch": 0.9723502304147466,
+      "grad_norm": 0.0,
+      "learning_rate": 9.135600163867268e-06,
+      "loss": 2.3196,
+      "step": 422
+    },
+    {
+      "epoch": 0.9746543778801844,
+      "grad_norm": 0.0,
+      "learning_rate": 9.131951569874822e-06,
+      "loss": 2.2138,
+      "step": 423
+    },
+    {
+      "epoch": 0.9769585253456221,
+      "grad_norm": 0.0,
+      "learning_rate": 9.128289473684213e-06,
+      "loss": 2.3115,
+      "step": 424
+    },
+    {
+      "epoch": 0.9792626728110599,
+      "grad_norm": 0.0,
+      "learning_rate": 9.124613800205974e-06,
+      "loss": 2.3028,
+      "step": 425
+    },
+    {
+      "epoch": 0.9815668202764977,
+      "grad_norm": 0.0,
+      "learning_rate": 9.120924473792819e-06,
+      "loss": 2.3505,
+      "step": 426
+    },
+    {
+      "epoch": 0.9838709677419355,
+      "grad_norm": 0.0,
+      "learning_rate": 9.117221418234443e-06,
+      "loss": 2.2614,
+      "step": 427
+    },
+    {
+      "epoch": 0.9861751152073732,
+      "grad_norm": 0.0,
+      "learning_rate": 9.113504556752279e-06,
+      "loss": 2.0659,
+      "step": 428
+    },
+    {
+      "epoch": 0.988479262672811,
+      "grad_norm": 0.0,
+      "learning_rate": 9.10977381199419e-06,
+      "loss": 2.0262,
+      "step": 429
+    },
+    {
+      "epoch": 0.9907834101382489,
+      "grad_norm": 0.0,
+      "learning_rate": 9.106029106029106e-06,
+      "loss": 2.4228,
+      "step": 430
+    },
+    {
+      "epoch": 0.9930875576036866,
+      "grad_norm": 0.0,
+      "learning_rate": 9.102270360341596e-06,
+      "loss": 1.9768,
+      "step": 431
+    },
+    {
+      "epoch": 0.9953917050691244,
+      "grad_norm": 0.0,
+      "learning_rate": 9.098497495826378e-06,
+      "loss": 1.817,
+      "step": 432
+    },
+    {
+      "epoch": 0.9976958525345622,
+      "grad_norm": 0.0,
+      "learning_rate": 9.094710432782772e-06,
+      "loss": 2.1856,
+      "step": 433
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 2.3401,
+      "step": 434
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 868,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 217,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0952722129291837e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fcb11f3fd40a8cc68b7b0f1c083ca59af623554766eefd61bd91def15fa28d31
+size 6968