First Commit

config.json

@@ -0,0 +1,23 @@
+{
+  "_name_or_path": "/scalar2pool/shared/FastChat/model_weights/psmathur_orca_mini_3b",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 3200,
+  "initializer_range": 0.02,
+  "intermediate_size": 8640,
+  "max_position_embeddings": 2048,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 26,
+  "pad_token_id": 0,
+  "rms_norm_eps": 1e-06,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.28.1",
+  "use_cache": false,
+  "vocab_size": 32000
+}

generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "pad_token_id": 0,
+  "transformers_version": "4.28.1"
+}

pytorch_model-00001-of-00002.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40d41e4f7580db32660ba5ddd73db076f38aa838dc4bb6efb5596d22fb449bdb
+size 9990701755

pytorch_model-00002-of-00002.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f72ae29e98f37e36dd5282b8e53fddafa295501c62f675b01facd8b055b29e46
+size 3715287557

pytorch_model.bin.index.json

@@ -0,0 +1,270 @@
+{
+  "metadata": {
+    "total_size": 13705899600
+  },
+  "weight_map": {
+    "lm_head.weight": "pytorch_model-00002-of-00002.bin",
+    "model.embed_tokens.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.20.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.rotary_emb.inv_freq": "pytorch_model-00002-of-00002.bin",
+    "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.rotary_emb.inv_freq": "pytorch_model-00002-of-00002.bin",
+    "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.rotary_emb.inv_freq": "pytorch_model-00002-of-00002.bin",
+    "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.rotary_emb.inv_freq": "pytorch_model-00002-of-00002.bin",
+    "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.rotary_emb.inv_freq": "pytorch_model-00002-of-00002.bin",
+    "model.layers.24.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.rotary_emb.inv_freq": "pytorch_model-00002-of-00002.bin",
+    "model.layers.25.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
+    "model.layers.3.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
+    "model.norm.weight": "pytorch_model-00002-of-00002.bin"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,12 @@
+{
+  "bos_token": "<s>",
+  "eos_token": "</s>",
+  "pad_token": "<unk>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab1b681ec7fc02fed5edd3026687d7a692a918c4dd8e150ca2e3994a6229843b
+size 534194

tokenizer_config.json

@@ -0,0 +1,34 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "bos_token": {
+    "__type": "AddedToken",
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "clean_up_tokenization_spaces": false,
+  "eos_token": {
+    "__type": "AddedToken",
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "model_max_length": 2048,
+  "pad_token": null,
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": {
+    "__type": "AddedToken",
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

trainer_state.json

@@ -0,0 +1,3391 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9959946595460614,
+  "global_step": 561,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 1.5677,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 1.5543,
+      "step": 2
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 1.4947,
+      "step": 3
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 1.2593,
+      "step": 4
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 1.163,
+      "step": 5
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.058823529411766e-06,
+      "loss": 1.1346,
+      "step": 6
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 1.052,
+      "step": 7
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.937,
+      "step": 8
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0588235294117648e-05,
+      "loss": 0.9712,
+      "step": 9
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 0.8918,
+      "step": 10
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2941176470588238e-05,
+      "loss": 0.8973,
+      "step": 11
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.4117647058823532e-05,
+      "loss": 0.8255,
+      "step": 12
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5294117647058822e-05,
+      "loss": 0.8799,
+      "step": 13
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.647058823529412e-05,
+      "loss": 0.7789,
+      "step": 14
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 0.8343,
+      "step": 15
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8823529411764708e-05,
+      "loss": 0.7583,
+      "step": 16
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 0.7006,
+      "step": 17
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999983324811832e-05,
+      "loss": 0.7948,
+      "step": 18
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999332998034515e-05,
+      "loss": 0.7367,
+      "step": 19
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999849926643211e-05,
+      "loss": 0.754,
+      "step": 20
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997332081116374e-05,
+      "loss": 0.7247,
+      "step": 21
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9995831481013376e-05,
+      "loss": 0.7246,
+      "step": 22
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999399751616869e-05,
+      "loss": 0.7369,
+      "step": 23
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9991830247745732e-05,
+      "loss": 0.6449,
+      "step": 24
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9989329748023728e-05,
+      "loss": 0.6778,
+      "step": 25
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9986496100395276e-05,
+      "loss": 0.6913,
+      "step": 26
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.99833293993636e-05,
+      "loss": 0.6488,
+      "step": 27
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.997982975053936e-05,
+      "loss": 0.66,
+      "step": 28
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9975997270637172e-05,
+      "loss": 0.6721,
+      "step": 29
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9971832087471678e-05,
+      "loss": 0.6535,
+      "step": 30
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9967334339953303e-05,
+      "loss": 0.6463,
+      "step": 31
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996250417808362e-05,
+      "loss": 0.6506,
+      "step": 32
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9957341762950346e-05,
+      "loss": 0.6617,
+      "step": 33
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.995184726672197e-05,
+      "loss": 0.6229,
+      "step": 34
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.994602087264201e-05,
+      "loss": 0.6246,
+      "step": 35
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9939862775022893e-05,
+      "loss": 0.5902,
+      "step": 36
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.99333731792395e-05,
+      "loss": 0.5928,
+      "step": 37
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992655230172229e-05,
+      "loss": 0.625,
+      "step": 38
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.99194003699501e-05,
+      "loss": 0.5902,
+      "step": 39
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.991191762244254e-05,
+      "loss": 0.5957,
+      "step": 40
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9904104308752053e-05,
+      "loss": 0.6345,
+      "step": 41
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9895960689455598e-05,
+      "loss": 0.6346,
+      "step": 42
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9887487036145942e-05,
+      "loss": 0.5765,
+      "step": 43
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9878683631422605e-05,
+      "loss": 0.5625,
+      "step": 44
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.986955076888246e-05,
+      "loss": 0.5533,
+      "step": 45
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9860088753109896e-05,
+      "loss": 0.5329,
+      "step": 46
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985029789966671e-05,
+      "loss": 0.6103,
+      "step": 47
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9840178535081548e-05,
+      "loss": 0.5537,
+      "step": 48
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.982973099683902e-05,
+      "loss": 0.5759,
+      "step": 49
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9818955633368464e-05,
+      "loss": 0.528,
+      "step": 50
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9807852804032306e-05,
+      "loss": 0.5689,
+      "step": 51
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9796422879114082e-05,
+      "loss": 0.5433,
+      "step": 52
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.978466623980609e-05,
+      "loss": 0.5588,
+      "step": 53
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9772583278196677e-05,
+      "loss": 0.5959,
+      "step": 54
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9760174397257158e-05,
+      "loss": 0.5158,
+      "step": 55
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9747440010828384e-05,
+      "loss": 0.5334,
+      "step": 56
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9734380543606932e-05,
+      "loss": 0.536,
+      "step": 57
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9720996431130946e-05,
+      "loss": 0.5678,
+      "step": 58
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9707288119765625e-05,
+      "loss": 0.5527,
+      "step": 59
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.96932560666883e-05,
+      "loss": 0.5685,
+      "step": 60
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.967890073987323e-05,
+      "loss": 0.5417,
+      "step": 61
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9664222618075958e-05,
+      "loss": 0.5456,
+      "step": 62
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9649222190817382e-05,
+      "loss": 0.5382,
+      "step": 63
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9633899958367384e-05,
+      "loss": 0.5357,
+      "step": 64
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.961825643172819e-05,
+      "loss": 0.5587,
+      "step": 65
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.96022921326173e-05,
+      "loss": 0.5525,
+      "step": 66
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9586007593450098e-05,
+      "loss": 0.5559,
+      "step": 67
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.956940335732209e-05,
+      "loss": 0.4953,
+      "step": 68
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9552479977990802e-05,
+      "loss": 0.5843,
+      "step": 69
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.95352380198573e-05,
+      "loss": 0.5403,
+      "step": 70
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9517678057947385e-05,
+      "loss": 0.5343,
+      "step": 71
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9499800677892386e-05,
+      "loss": 0.57,
+      "step": 72
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.948160647590966e-05,
+      "loss": 0.5127,
+      "step": 73
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.946309605878269e-05,
+      "loss": 0.5099,
+      "step": 74
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9444270043840854e-05,
+      "loss": 0.5108,
+      "step": 75
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9425129058938833e-05,
+      "loss": 0.5676,
+      "step": 76
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9405673742435677e-05,
+      "loss": 0.4828,
+      "step": 77
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.938590474317352e-05,
+      "loss": 0.5479,
+      "step": 78
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9365822720455915e-05,
+      "loss": 0.5083,
+      "step": 79
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9345428344025883e-05,
+      "loss": 0.558,
+      "step": 80
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.932472229404356e-05,
+      "loss": 0.5583,
+      "step": 81
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9303705261063496e-05,
+      "loss": 0.5335,
+      "step": 82
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.928237794601165e-05,
+      "loss": 0.523,
+      "step": 83
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9260741060162015e-05,
+      "loss": 0.5088,
+      "step": 84
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9238795325112867e-05,
+      "loss": 0.4811,
+      "step": 85
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9216541472762736e-05,
+      "loss": 0.4711,
+      "step": 86
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9193980245285967e-05,
+      "loss": 0.5165,
+      "step": 87
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9171112395107988e-05,
+      "loss": 0.5411,
+      "step": 88
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9147938684880213e-05,
+      "loss": 0.5248,
+      "step": 89
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.912445988745459e-05,
+      "loss": 0.4863,
+      "step": 90
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9100676785857862e-05,
+      "loss": 0.4909,
+      "step": 91
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9076590173265406e-05,
+      "loss": 0.4965,
+      "step": 92
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.905220085297482e-05,
+      "loss": 0.5159,
+      "step": 93
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9027509638379122e-05,
+      "loss": 0.5308,
+      "step": 94
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.90025173529396e-05,
+      "loss": 0.4859,
+      "step": 95
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.897722483015838e-05,
+      "loss": 0.482,
+      "step": 96
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8951632913550625e-05,
+      "loss": 0.51,
+      "step": 97
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8925742456616375e-05,
+      "loss": 0.4939,
+      "step": 98
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.889955432281212e-05,
+      "loss": 0.473,
+      "step": 99
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.887306938552197e-05,
+      "loss": 0.5204,
+      "step": 100
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8846288528028555e-05,
+      "loss": 0.531,
+      "step": 101
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.881921264348355e-05,
+      "loss": 0.5008,
+      "step": 102
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.87918426348779e-05,
+      "loss": 0.5119,
+      "step": 103
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.876417941501168e-05,
+      "loss": 0.4895,
+      "step": 104
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8736223906463698e-05,
+      "loss": 0.4326,
+      "step": 105
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8707977041560673e-05,
+      "loss": 0.5274,
+      "step": 106
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8679439762346186e-05,
+      "loss": 0.4578,
+      "step": 107
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8650613020549232e-05,
+      "loss": 0.4559,
+      "step": 108
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8621497777552508e-05,
+      "loss": 0.5023,
+      "step": 109
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8592095004360316e-05,
+      "loss": 0.5008,
+      "step": 110
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8562405681566217e-05,
+      "loss": 0.4788,
+      "step": 111
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.853243079932029e-05,
+      "loss": 0.4943,
+      "step": 112
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8502171357296144e-05,
+      "loss": 0.4343,
+      "step": 113
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8471628364657555e-05,
+      "loss": 0.4892,
+      "step": 114
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8440802840024824e-05,
+      "loss": 0.4059,
+      "step": 115
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8409695811440796e-05,
+      "loss": 0.4538,
+      "step": 116
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8378308316336585e-05,
+      "loss": 0.4747,
+      "step": 117
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.834664140149696e-05,
+      "loss": 0.4832,
+      "step": 118
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8314696123025456e-05,
+      "loss": 0.4787,
+      "step": 119
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.828247354630912e-05,
+      "loss": 0.424,
+      "step": 120
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8249974745983023e-05,
+      "loss": 0.5097,
+      "step": 121
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8217200805894382e-05,
+      "loss": 0.4866,
+      "step": 122
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8184152819066437e-05,
+      "loss": 0.4566,
+      "step": 123
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.815083188766198e-05,
+      "loss": 0.482,
+      "step": 124
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8117239122946615e-05,
+      "loss": 0.4329,
+      "step": 125
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8083375645251687e-05,
+      "loss": 0.4572,
+      "step": 126
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8049242583936923e-05,
+      "loss": 0.4645,
+      "step": 127
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8014841077352764e-05,
+      "loss": 0.4544,
+      "step": 128
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7980172272802398e-05,
+      "loss": 0.5108,
+      "step": 129
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7945237326503507e-05,
+      "loss": 0.4418,
+      "step": 130
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7910037403549695e-05,
+      "loss": 0.4852,
+      "step": 131
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7874573677871638e-05,
+      "loss": 0.4361,
+      "step": 132
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.783884733219794e-05,
+      "loss": 0.4399,
+      "step": 133
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7802859558015666e-05,
+      "loss": 0.4763,
+      "step": 134
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7766611555530638e-05,
+      "loss": 0.4876,
+      "step": 135
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.773010453362737e-05,
+      "loss": 0.4602,
+      "step": 136
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.769333970982879e-05,
+      "loss": 0.4436,
+      "step": 137
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7656318310255604e-05,
+      "loss": 0.4583,
+      "step": 138
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.761904156958542e-05,
+      "loss": 0.503,
+      "step": 139
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.758151073101157e-05,
+      "loss": 0.4634,
+      "step": 140
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7543727046201642e-05,
+      "loss": 0.4806,
+      "step": 141
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7505691775255744e-05,
+      "loss": 0.4714,
+      "step": 142
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7467406186664474e-05,
+      "loss": 0.4972,
+      "step": 143
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7428871557266628e-05,
+      "loss": 0.4341,
+      "step": 144
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7390089172206594e-05,
+      "loss": 0.4768,
+      "step": 145
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7351060324891506e-05,
+      "loss": 0.476,
+      "step": 146
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7311786316948112e-05,
+      "loss": 0.4397,
+      "step": 147
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7272268458179352e-05,
+      "loss": 0.4453,
+      "step": 148
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7232508066520702e-05,
+      "loss": 0.4612,
+      "step": 149
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7192506467996174e-05,
+      "loss": 0.428,
+      "step": 150
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7152264996674138e-05,
+      "loss": 0.4876,
+      "step": 151
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7111784994622804e-05,
+      "loss": 0.4717,
+      "step": 152
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.4537,
+      "step": 153
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7030114806335528e-05,
+      "loss": 0.423,
+      "step": 154
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6988927343831093e-05,
+      "loss": 0.4351,
+      "step": 155
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6947506797969563e-05,
+      "loss": 0.4287,
+      "step": 156
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6905854550141717e-05,
+      "loss": 0.4379,
+      "step": 157
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.68639719894657e-05,
+      "loss": 0.4335,
+      "step": 158
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6821860512740674e-05,
+      "loss": 0.4726,
+      "step": 159
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6779521524400234e-05,
+      "loss": 0.4366,
+      "step": 160
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6736956436465573e-05,
+      "loss": 0.4582,
+      "step": 161
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6694166668498396e-05,
+      "loss": 0.4766,
+      "step": 162
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.665115364755357e-05,
+      "loss": 0.465,
+      "step": 163
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6607918808131526e-05,
+      "loss": 0.4289,
+      "step": 164
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.656446359213043e-05,
+      "loss": 0.4643,
+      "step": 165
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6520789448798086e-05,
+      "loss": 0.4125,
+      "step": 166
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.647689783468362e-05,
+      "loss": 0.4837,
+      "step": 167
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6432790213588874e-05,
+      "loss": 0.4359,
+      "step": 168
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.638846805651961e-05,
+      "loss": 0.4446,
+      "step": 169
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6343932841636455e-05,
+      "loss": 0.404,
+      "step": 170
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.629918605420558e-05,
+      "loss": 0.4411,
+      "step": 171
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.625422918654918e-05,
+      "loss": 0.4658,
+      "step": 172
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6209063737995716e-05,
+      "loss": 0.443,
+      "step": 173
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6163691214829895e-05,
+      "loss": 0.4379,
+      "step": 174
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6118113130242435e-05,
+      "loss": 0.4624,
+      "step": 175
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6072331004279617e-05,
+      "loss": 0.4448,
+      "step": 176
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6026346363792565e-05,
+      "loss": 0.4381,
+      "step": 177
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.598016074238635e-05,
+      "loss": 0.4679,
+      "step": 178
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5933775680368825e-05,
+      "loss": 0.4162,
+      "step": 179
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5887192724699263e-05,
+      "loss": 0.4401,
+      "step": 180
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5840413428936767e-05,
+      "loss": 0.4354,
+      "step": 181
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.579343935318846e-05,
+      "loss": 0.4898,
+      "step": 182
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5746272064057438e-05,
+      "loss": 0.4489,
+      "step": 183
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5698913134590552e-05,
+      "loss": 0.441,
+      "step": 184
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.565136414422592e-05,
+      "loss": 0.4295,
+      "step": 185
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5603626678740266e-05,
+      "loss": 0.4699,
+      "step": 186
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5555702330196024e-05,
+      "loss": 0.4161,
+      "step": 187
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5507592696888258e-05,
+      "loss": 0.4164,
+      "step": 188
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5459299383291347e-05,
+      "loss": 0.3109,
+      "step": 189
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.541082400000547e-05,
+      "loss": 0.3426,
+      "step": 190
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.53621681637029e-05,
+      "loss": 0.3539,
+      "step": 191
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5313333497074094e-05,
+      "loss": 0.3434,
+      "step": 192
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.526432162877356e-05,
+      "loss": 0.3437,
+      "step": 193
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.521513419336555e-05,
+      "loss": 0.337,
+      "step": 194
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5165772831269547e-05,
+      "loss": 0.3605,
+      "step": 195
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5116239188705557e-05,
+      "loss": 0.3378,
+      "step": 196
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5066534917639195e-05,
+      "loss": 0.341,
+      "step": 197
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.501666167572661e-05,
+      "loss": 0.3343,
+      "step": 198
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4966621126259184e-05,
+      "loss": 0.3791,
+      "step": 199
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.491641493810808e-05,
+      "loss": 0.3072,
+      "step": 200
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4866044785668563e-05,
+      "loss": 0.3812,
+      "step": 201
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4815512348804177e-05,
+      "loss": 0.3664,
+      "step": 202
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4764819312790706e-05,
+      "loss": 0.3283,
+      "step": 203
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4713967368259981e-05,
+      "loss": 0.3594,
+      "step": 204
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.466295821114348e-05,
+      "loss": 0.3653,
+      "step": 205
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4611793542615805e-05,
+      "loss": 0.3525,
+      "step": 206
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4560475069037895e-05,
+      "loss": 0.3384,
+      "step": 207
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4509004501900161e-05,
+      "loss": 0.3211,
+      "step": 208
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4457383557765385e-05,
+      "loss": 0.3125,
+      "step": 209
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4405613958211482e-05,
+      "loss": 0.3541,
+      "step": 210
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4353697429774083e-05,
+      "loss": 0.331,
+      "step": 211
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4301635703888946e-05,
+      "loss": 0.3354,
+      "step": 212
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4249430516834222e-05,
+      "loss": 0.3258,
+      "step": 213
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4197083609672543e-05,
+      "loss": 0.3267,
+      "step": 214
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4144596728192972e-05,
+      "loss": 0.3189,
+      "step": 215
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4091971622852751e-05,
+      "loss": 0.3343,
+      "step": 216
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.403921004871895e-05,
+      "loss": 0.3405,
+      "step": 217
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3986313765409924e-05,
+      "loss": 0.3095,
+      "step": 218
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3933284537036626e-05,
+      "loss": 0.3699,
+      "step": 219
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3880124132143782e-05,
+      "loss": 0.3497,
+      "step": 220
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3826834323650899e-05,
+      "loss": 0.379,
+      "step": 221
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3773416888793145e-05,
+      "loss": 0.3415,
+      "step": 222
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3719873609062078e-05,
+      "loss": 0.3399,
+      "step": 223
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3666206270146223e-05,
+      "loss": 0.3307,
+      "step": 224
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3612416661871532e-05,
+      "loss": 0.3774,
+      "step": 225
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3558506578141683e-05,
+      "loss": 0.3684,
+      "step": 226
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.350447781687826e-05,
+      "loss": 0.3617,
+      "step": 227
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.345033217996078e-05,
+      "loss": 0.3129,
+      "step": 228
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3396071473166614e-05,
+      "loss": 0.3475,
+      "step": 229
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3341697506110753e-05,
+      "loss": 0.3387,
+      "step": 230
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3287212092185464e-05,
+      "loss": 0.2923,
+      "step": 231
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3232617048499801e-05,
+      "loss": 0.3306,
+      "step": 232
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3177914195819018e-05,
+      "loss": 0.341,
+      "step": 233
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3123105358503839e-05,
+      "loss": 0.3279,
+      "step": 234
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3068192364449618e-05,
+      "loss": 0.3457,
+      "step": 235
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3013177045025374e-05,
+      "loss": 0.3409,
+      "step": 236
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2958061235012707e-05,
+      "loss": 0.3242,
+      "step": 237
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2902846772544625e-05,
+      "loss": 0.3319,
+      "step": 238
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2847535499044232e-05,
+      "loss": 0.3254,
+      "step": 239
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.279212925916332e-05,
+      "loss": 0.3246,
+      "step": 240
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2736629900720832e-05,
+      "loss": 0.34,
+      "step": 241
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2681039274641261e-05,
+      "loss": 0.3249,
+      "step": 242
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2625359234892906e-05,
+      "loss": 0.3063,
+      "step": 243
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2569591638426054e-05,
+      "loss": 0.3374,
+      "step": 244
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2513738345111029e-05,
+      "loss": 0.332,
+      "step": 245
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2457801217676182e-05,
+      "loss": 0.3159,
+      "step": 246
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2401782121645767e-05,
+      "loss": 0.3317,
+      "step": 247
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2345682925277716e-05,
+      "loss": 0.3115,
+      "step": 248
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2289505499501341e-05,
+      "loss": 0.3331,
+      "step": 249
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2233251717854937e-05,
+      "loss": 0.3081,
+      "step": 250
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2176923456423283e-05,
+      "loss": 0.3021,
+      "step": 251
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2120522593775108e-05,
+      "loss": 0.3608,
+      "step": 252
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2064051010900397e-05,
+      "loss": 0.3622,
+      "step": 253
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2007510591147698e-05,
+      "loss": 0.2896,
+      "step": 254
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1950903220161286e-05,
+      "loss": 0.3336,
+      "step": 255
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1894230785818284e-05,
+      "loss": 0.3206,
+      "step": 256
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1837495178165706e-05,
+      "loss": 0.3196,
+      "step": 257
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1780698289357419e-05,
+      "loss": 0.3602,
+      "step": 258
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1723842013591044e-05,
+      "loss": 0.3331,
+      "step": 259
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1666928247044769e-05,
+      "loss": 0.3131,
+      "step": 260
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.160995888781413e-05,
+      "loss": 0.3215,
+      "step": 261
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1552935835848697e-05,
+      "loss": 0.3626,
+      "step": 262
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1495860992888712e-05,
+      "loss": 0.3606,
+      "step": 263
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1438736262401669e-05,
+      "loss": 0.2954,
+      "step": 264
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1381563549518823e-05,
+      "loss": 0.3466,
+      "step": 265
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.132434476097167e-05,
+      "loss": 0.3221,
+      "step": 266
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.126708180502834e-05,
+      "loss": 0.3279,
+      "step": 267
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.120977659142996e-05,
+      "loss": 0.2793,
+      "step": 268
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1152431031326978e-05,
+      "loss": 0.3143,
+      "step": 269
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1095047037215397e-05,
+      "loss": 0.3365,
+      "step": 270
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.103762652287302e-05,
+      "loss": 0.3223,
+      "step": 271
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.098017140329561e-05,
+      "loss": 0.3093,
+      "step": 272
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.092268359463302e-05,
+      "loss": 0.2947,
+      "step": 273
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0865165014125316e-05,
+      "loss": 0.27,
+      "step": 274
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0807617580038797e-05,
+      "loss": 0.2949,
+      "step": 275
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0750043211602045e-05,
+      "loss": 0.3023,
+      "step": 276
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0692443828941918e-05,
+      "loss": 0.3187,
+      "step": 277
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0634821353019505e-05,
+      "loss": 0.319,
+      "step": 278
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0577177705566061e-05,
+      "loss": 0.3289,
+      "step": 279
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0519514809018927e-05,
+      "loss": 0.3083,
+      "step": 280
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0461834586457398e-05,
+      "loss": 0.2893,
+      "step": 281
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0404138961538603e-05,
+      "loss": 0.2896,
+      "step": 282
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0346429858433354e-05,
+      "loss": 0.3484,
+      "step": 283
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0288709201761949e-05,
+      "loss": 0.319,
+      "step": 284
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0230978916530012e-05,
+      "loss": 0.3459,
+      "step": 285
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0173240928064285e-05,
+      "loss": 0.3632,
+      "step": 286
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0115497161948409e-05,
+      "loss": 0.2869,
+      "step": 287
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0057749543958717e-05,
+      "loss": 0.3111,
+      "step": 288
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3112,
+      "step": 289
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.942250456041286e-06,
+      "loss": 0.3453,
+      "step": 290
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.884502838051595e-06,
+      "loss": 0.3241,
+      "step": 291
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.826759071935718e-06,
+      "loss": 0.3212,
+      "step": 292
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.769021083469991e-06,
+      "loss": 0.3331,
+      "step": 293
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.711290798238056e-06,
+      "loss": 0.3268,
+      "step": 294
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.653570141566653e-06,
+      "loss": 0.3517,
+      "step": 295
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.595861038461399e-06,
+      "loss": 0.2879,
+      "step": 296
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.538165413542607e-06,
+      "loss": 0.3023,
+      "step": 297
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.480485190981073e-06,
+      "loss": 0.2807,
+      "step": 298
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.422822294433939e-06,
+      "loss": 0.282,
+      "step": 299
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.365178646980497e-06,
+      "loss": 0.3262,
+      "step": 300
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.307556171058085e-06,
+      "loss": 0.3152,
+      "step": 301
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.249956788397956e-06,
+      "loss": 0.3316,
+      "step": 302
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.192382419961208e-06,
+      "loss": 0.3172,
+      "step": 303
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.134834985874687e-06,
+      "loss": 0.289,
+      "step": 304
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.07731640536698e-06,
+      "loss": 0.3397,
+      "step": 305
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.019828596704394e-06,
+      "loss": 0.3056,
+      "step": 306
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.962373477126983e-06,
+      "loss": 0.3063,
+      "step": 307
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.904952962784605e-06,
+      "loss": 0.3214,
+      "step": 308
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.847568968673025e-06,
+      "loss": 0.313,
+      "step": 309
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.790223408570043e-06,
+      "loss": 0.3421,
+      "step": 310
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.732918194971663e-06,
+      "loss": 0.313,
+      "step": 311
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.675655239028333e-06,
+      "loss": 0.3239,
+      "step": 312
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.618436450481182e-06,
+      "loss": 0.3127,
+      "step": 313
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.561263737598338e-06,
+      "loss": 0.3185,
+      "step": 314
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.50413900711129e-06,
+      "loss": 0.3184,
+      "step": 315
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.447064164151305e-06,
+      "loss": 0.3155,
+      "step": 316
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.39004111218587e-06,
+      "loss": 0.348,
+      "step": 317
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.333071752955233e-06,
+      "loss": 0.3455,
+      "step": 318
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.276157986408959e-06,
+      "loss": 0.3466,
+      "step": 319
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.219301710642583e-06,
+      "loss": 0.2885,
+      "step": 320
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.162504821834296e-06,
+      "loss": 0.3207,
+      "step": 321
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.10576921418172e-06,
+      "loss": 0.3284,
+      "step": 322
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.04909677983872e-06,
+      "loss": 0.3168,
+      "step": 323
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.992489408852306e-06,
+      "loss": 0.2769,
+      "step": 324
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.935948989099606e-06,
+      "loss": 0.2779,
+      "step": 325
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.879477406224894e-06,
+      "loss": 0.3447,
+      "step": 326
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.823076543576718e-06,
+      "loss": 0.3574,
+      "step": 327
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.766748282145068e-06,
+      "loss": 0.2785,
+      "step": 328
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.710494500498662e-06,
+      "loss": 0.318,
+      "step": 329
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.654317074722287e-06,
+      "loss": 0.3426,
+      "step": 330
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.598217878354237e-06,
+      "loss": 0.3269,
+      "step": 331
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.542198782323819e-06,
+      "loss": 0.3034,
+      "step": 332
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.486261654888974e-06,
+      "loss": 0.3239,
+      "step": 333
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.430408361573949e-06,
+      "loss": 0.3193,
+      "step": 334
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.374640765107095e-06,
+      "loss": 0.3178,
+      "step": 335
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.318960725358742e-06,
+      "loss": 0.3133,
+      "step": 336
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.263370099279173e-06,
+      "loss": 0.3128,
+      "step": 337
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.207870740836684e-06,
+      "loss": 0.2936,
+      "step": 338
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.152464500955769e-06,
+      "loss": 0.3435,
+      "step": 339
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.097153227455379e-06,
+      "loss": 0.2857,
+      "step": 340
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.041938764987297e-06,
+      "loss": 0.3168,
+      "step": 341
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.986822954974631e-06,
+      "loss": 0.3354,
+      "step": 342
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.9318076355503835e-06,
+      "loss": 0.3187,
+      "step": 343
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.876894641496164e-06,
+      "loss": 0.3088,
+      "step": 344
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.822085804180985e-06,
+      "loss": 0.3193,
+      "step": 345
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.767382951500205e-06,
+      "loss": 0.3283,
+      "step": 346
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.712787907814542e-06,
+      "loss": 0.3457,
+      "step": 347
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.658302493889251e-06,
+      "loss": 0.3085,
+      "step": 348
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.603928526833386e-06,
+      "loss": 0.3032,
+      "step": 349
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.549667820039221e-06,
+      "loss": 0.301,
+      "step": 350
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.495522183121741e-06,
+      "loss": 0.3014,
+      "step": 351
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.441493421858318e-06,
+      "loss": 0.3218,
+      "step": 352
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.387583338128471e-06,
+      "loss": 0.2764,
+      "step": 353
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.333793729853781e-06,
+      "loss": 0.3088,
+      "step": 354
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.280126390937925e-06,
+      "loss": 0.2784,
+      "step": 355
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.2265831112068565e-06,
+      "loss": 0.3194,
+      "step": 356
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.173165676349103e-06,
+      "loss": 0.3269,
+      "step": 357
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.11987586785622e-06,
+      "loss": 0.3102,
+      "step": 358
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.0667154629633766e-06,
+      "loss": 0.2983,
+      "step": 359
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.013686234590077e-06,
+      "loss": 0.2933,
+      "step": 360
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.960789951281052e-06,
+      "loss": 0.3244,
+      "step": 361
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.908028377147252e-06,
+      "loss": 0.3312,
+      "step": 362
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.855403271807033e-06,
+      "loss": 0.2832,
+      "step": 363
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.802916390327459e-06,
+      "loss": 0.3005,
+      "step": 364
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.750569483165785e-06,
+      "loss": 0.2536,
+      "step": 365
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.698364296111057e-06,
+      "loss": 0.2949,
+      "step": 366
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.646302570225919e-06,
+      "loss": 0.3253,
+      "step": 367
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.59438604178852e-06,
+      "loss": 0.294,
+      "step": 368
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.542616442234618e-06,
+      "loss": 0.3566,
+      "step": 369
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.490995498099844e-06,
+      "loss": 0.3137,
+      "step": 370
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.43952493096211e-06,
+      "loss": 0.3113,
+      "step": 371
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.388206457384198e-06,
+      "loss": 0.2576,
+      "step": 372
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.337041788856518e-06,
+      "loss": 0.3104,
+      "step": 373
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.286032631740023e-06,
+      "loss": 0.29,
+      "step": 374
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.235180687209296e-06,
+      "loss": 0.276,
+      "step": 375
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.184487651195825e-06,
+      "loss": 0.2317,
+      "step": 376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.133955214331439e-06,
+      "loss": 0.2259,
+      "step": 377
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.083585061891925e-06,
+      "loss": 0.2627,
+      "step": 378
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.03337887374082e-06,
+      "loss": 0.2296,
+      "step": 379
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.983338324273397e-06,
+      "loss": 0.2251,
+      "step": 380
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.933465082360808e-06,
+      "loss": 0.2308,
+      "step": 381
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.8837608112944456e-06,
+      "loss": 0.2311,
+      "step": 382
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.834227168730451e-06,
+      "loss": 0.2189,
+      "step": 383
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.784865806634449e-06,
+      "loss": 0.1893,
+      "step": 384
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.7356783712264405e-06,
+      "loss": 0.2325,
+      "step": 385
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.686666502925908e-06,
+      "loss": 0.2168,
+      "step": 386
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.637831836297103e-06,
+      "loss": 0.233,
+      "step": 387
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.589175999994535e-06,
+      "loss": 0.2409,
+      "step": 388
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.5407006167086575e-06,
+      "loss": 0.23,
+      "step": 389
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.492407303111745e-06,
+      "loss": 0.2304,
+      "step": 390
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.444297669803981e-06,
+      "loss": 0.2168,
+      "step": 391
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.396373321259737e-06,
+      "loss": 0.2405,
+      "step": 392
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.348635855774082e-06,
+      "loss": 0.2307,
+      "step": 393
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.301086865409449e-06,
+      "loss": 0.2167,
+      "step": 394
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.253727935942563e-06,
+      "loss": 0.2145,
+      "step": 395
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.206560646811545e-06,
+      "loss": 0.2379,
+      "step": 396
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.1595865710632366e-06,
+      "loss": 0.2261,
+      "step": 397
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.112807275300742e-06,
+      "loss": 0.2392,
+      "step": 398
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.066224319631181e-06,
+      "loss": 0.2267,
+      "step": 399
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.019839257613652e-06,
+      "loss": 0.217,
+      "step": 400
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.973653636207437e-06,
+      "loss": 0.2074,
+      "step": 401
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.927668995720384e-06,
+      "loss": 0.2611,
+      "step": 402
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.881886869757565e-06,
+      "loss": 0.2489,
+      "step": 403
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.836308785170109e-06,
+      "loss": 0.2227,
+      "step": 404
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.790936262004287e-06,
+      "loss": 0.24,
+      "step": 405
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.745770813450824e-06,
+      "loss": 0.2136,
+      "step": 406
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.700813945794425e-06,
+      "loss": 0.221,
+      "step": 407
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6560671583635467e-06,
+      "loss": 0.2485,
+      "step": 408
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6115319434803897e-06,
+      "loss": 0.2369,
+      "step": 409
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.5672097864111287e-06,
+      "loss": 0.203,
+      "step": 410
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.523102165316381e-06,
+      "loss": 0.2205,
+      "step": 411
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.4792105512019148e-06,
+      "loss": 0.2016,
+      "step": 412
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.435536407869575e-06,
+      "loss": 0.2217,
+      "step": 413
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.3920811918684804e-06,
+      "loss": 0.2247,
+      "step": 414
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3488463524464355e-06,
+      "loss": 0.234,
+      "step": 415
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3058333315016066e-06,
+      "loss": 0.224,
+      "step": 416
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2630435635344283e-06,
+      "loss": 0.2364,
+      "step": 417
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.22047847559977e-06,
+      "loss": 0.2006,
+      "step": 418
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1781394872593296e-06,
+      "loss": 0.2109,
+      "step": 419
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.136028010534303e-06,
+      "loss": 0.2188,
+      "step": 420
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.0941454498582847e-06,
+      "loss": 0.2723,
+      "step": 421
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.05249320203044e-06,
+      "loss": 0.2298,
+      "step": 422
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.011072656168906e-06,
+      "loss": 0.2503,
+      "step": 423
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.9698851936644767e-06,
+      "loss": 0.2141,
+      "step": 424
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.2047,
+      "step": 425
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8882150053771997e-06,
+      "loss": 0.2216,
+      "step": 426
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.847735003325868e-06,
+      "loss": 0.2178,
+      "step": 427
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.807493532003831e-06,
+      "loss": 0.2302,
+      "step": 428
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.767491933479304e-06,
+      "loss": 0.2353,
+      "step": 429
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.7277315418206476e-06,
+      "loss": 0.2275,
+      "step": 430
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.6882136830518923e-06,
+      "loss": 0.193,
+      "step": 431
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6489396751084983e-06,
+      "loss": 0.2215,
+      "step": 432
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6099108277934105e-06,
+      "loss": 0.2201,
+      "step": 433
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5711284427333716e-06,
+      "loss": 0.2439,
+      "step": 434
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.532593813335524e-06,
+      "loss": 0.2209,
+      "step": 435
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.4943082247442584e-06,
+      "loss": 0.2283,
+      "step": 436
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.456272953798361e-06,
+      "loss": 0.2175,
+      "step": 437
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.418489268988433e-06,
+      "loss": 0.2117,
+      "step": 438
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.3809584304145827e-06,
+      "loss": 0.2253,
+      "step": 439
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.343681689744396e-06,
+      "loss": 0.2502,
+      "step": 440
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.306660290171211e-06,
+      "loss": 0.2309,
+      "step": 441
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.26989546637263e-06,
+      "loss": 0.2318,
+      "step": 442
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2333884444693656e-06,
+      "loss": 0.2524,
+      "step": 443
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1971404419843355e-06,
+      "loss": 0.2109,
+      "step": 444
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1611526678020658e-06,
+      "loss": 0.2242,
+      "step": 445
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1254263221283657e-06,
+      "loss": 0.214,
+      "step": 446
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0899625964503113e-06,
+      "loss": 0.2276,
+      "step": 447
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0547626734965e-06,
+      "loss": 0.2172,
+      "step": 448
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.019827727197605e-06,
+      "loss": 0.2347,
+      "step": 449
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.9851589226472402e-06,
+      "loss": 0.2418,
+      "step": 450
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.950757416063077e-06,
+      "loss": 0.2229,
+      "step": 451
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9166243547483143e-06,
+      "loss": 0.1961,
+      "step": 452
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.882760877053388e-06,
+      "loss": 0.235,
+      "step": 453
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8491681123380235e-06,
+      "loss": 0.2038,
+      "step": 454
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8158471809335653e-06,
+      "loss": 0.234,
+      "step": 455
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7827991941056177e-06,
+      "loss": 0.2004,
+      "step": 456
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7500252540169782e-06,
+      "loss": 0.2263,
+      "step": 457
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.717526453690881e-06,
+      "loss": 0.2253,
+      "step": 458
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.6853038769745466e-06,
+      "loss": 0.2229,
+      "step": 459
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6533585985030398e-06,
+      "loss": 0.1952,
+      "step": 460
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6216916836634179e-06,
+      "loss": 0.2302,
+      "step": 461
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5903041885592052e-06,
+      "loss": 0.214,
+      "step": 462
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.55919715997518e-06,
+      "loss": 0.2203,
+      "step": 463
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5283716353424482e-06,
+      "loss": 0.2135,
+      "step": 464
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.4978286427038602e-06,
+      "loss": 0.2331,
+      "step": 465
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4675692006797137e-06,
+      "loss": 0.2322,
+      "step": 466
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4375943184337871e-06,
+      "loss": 0.2126,
+      "step": 467
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4079049956396828e-06,
+      "loss": 0.2355,
+      "step": 468
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.3785022224474943e-06,
+      "loss": 0.2175,
+      "step": 469
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3493869794507664e-06,
+      "loss": 0.2439,
+      "step": 470
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3205602376538162e-06,
+      "loss": 0.1999,
+      "step": 471
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.2920229584393284e-06,
+      "loss": 0.2498,
+      "step": 472
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2637760935363053e-06,
+      "loss": 0.2111,
+      "step": 473
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2358205849883197e-06,
+      "loss": 0.1923,
+      "step": 474
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2081573651221036e-06,
+      "loss": 0.2083,
+      "step": 475
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1807873565164507e-06,
+      "loss": 0.2374,
+      "step": 476
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1537114719714482e-06,
+      "loss": 0.2186,
+      "step": 477
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1269306144780335e-06,
+      "loss": 0.2166,
+      "step": 478
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1004456771878836e-06,
+      "loss": 0.2311,
+      "step": 479
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.0742575433836255e-06,
+      "loss": 0.2217,
+      "step": 480
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0483670864493777e-06,
+      "loss": 0.2085,
+      "step": 481
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.02277516984162e-06,
+      "loss": 0.2478,
+      "step": 482
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.97482647060405e-07,
+      "loss": 0.2432,
+      "step": 483
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.724903616208837e-07,
+      "loss": 0.226,
+      "step": 484
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.47799147025179e-07,
+      "loss": 0.2041,
+      "step": 485
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.234098267345959e-07,
+      "loss": 0.1953,
+      "step": 486
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 8.993232141421415e-07,
+      "loss": 0.2397,
+      "step": 487
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.75540112545411e-07,
+      "loss": 0.2154,
+      "step": 488
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.520613151197899e-07,
+      "loss": 0.2189,
+      "step": 489
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.288876048920125e-07,
+      "loss": 0.2073,
+      "step": 490
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.060197547140347e-07,
+      "loss": 0.216,
+      "step": 491
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.834585272372663e-07,
+      "loss": 0.2242,
+      "step": 492
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.612046748871327e-07,
+      "loss": 0.218,
+      "step": 493
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.392589398379868e-07,
+      "loss": 0.2554,
+      "step": 494
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.176220539883494e-07,
+      "loss": 0.2117,
+      "step": 495
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.962947389365071e-07,
+      "loss": 0.1955,
+      "step": 496
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.752777059564431e-07,
+      "loss": 0.2325,
+      "step": 497
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.545716559741166e-07,
+      "loss": 0.2465,
+      "step": 498
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.34177279544087e-07,
+      "loss": 0.2298,
+      "step": 499
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.140952568264858e-07,
+      "loss": 0.2046,
+      "step": 500
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.943262575643239e-07,
+      "loss": 0.203,
+      "step": 501
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.748709410611686e-07,
+      "loss": 0.2208,
+      "step": 502
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.557299561591478e-07,
+      "loss": 0.2193,
+      "step": 503
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.369039412173116e-07,
+      "loss": 0.1831,
+      "step": 504
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.183935240903415e-07,
+      "loss": 0.2274,
+      "step": 505
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.001993221076162e-07,
+      "loss": 0.1928,
+      "step": 506
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.823219420526182e-07,
+      "loss": 0.2306,
+      "step": 507
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.6476198014269945e-07,
+      "loss": 0.217,
+      "step": 508
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.475200220092002e-07,
+      "loss": 0.2531,
+      "step": 509
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.305966426779118e-07,
+      "loss": 0.2348,
+      "step": 510
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.139924065499035e-07,
+      "loss": 0.2273,
+      "step": 511
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 3.977078673826995e-07,
+      "loss": 0.2514,
+      "step": 512
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.817435682718096e-07,
+      "loss": 0.1732,
+      "step": 513
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.661000416326177e-07,
+      "loss": 0.2039,
+      "step": 514
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.5077780918262196e-07,
+      "loss": 0.2312,
+      "step": 515
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.3577738192404395e-07,
+      "loss": 0.2413,
+      "step": 516
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.2109926012677484e-07,
+      "loss": 0.2151,
+      "step": 517
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.067439333117028e-07,
+      "loss": 0.2072,
+      "step": 518
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.927118802343787e-07,
+      "loss": 0.1867,
+      "step": 519
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.79003568869054e-07,
+      "loss": 0.2205,
+      "step": 520
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.656194563930714e-07,
+      "loss": 0.2377,
+      "step": 521
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.5255998917161903e-07,
+      "loss": 0.2024,
+      "step": 522
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.398256027428436e-07,
+      "loss": 0.2358,
+      "step": 523
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2741672180332409e-07,
+      "loss": 0.2019,
+      "step": 524
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.1533376019391095e-07,
+      "loss": 0.1883,
+      "step": 525
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0357712088591942e-07,
+      "loss": 0.221,
+      "step": 526
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.921471959676957e-07,
+      "loss": 0.2271,
+      "step": 527
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.8104436663153757e-07,
+      "loss": 0.2211,
+      "step": 528
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.7026900316098217e-07,
+      "loss": 0.2241,
+      "step": 529
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.5982146491845596e-07,
+      "loss": 0.2087,
+      "step": 530
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4970210033329102e-07,
+      "loss": 0.2444,
+      "step": 531
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.3991124689010426e-07,
+      "loss": 0.1825,
+      "step": 532
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.3044923111754427e-07,
+      "loss": 0.2217,
+      "step": 533
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2131636857739548e-07,
+      "loss": 0.2277,
+      "step": 534
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.125129638540623e-07,
+      "loss": 0.2429,
+      "step": 535
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.0403931054440375e-07,
+      "loss": 0.175,
+      "step": 536
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.589569124794918e-08,
+      "loss": 0.2254,
+      "step": 537
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.808237755746352e-08,
+      "loss": 0.1987,
+      "step": 538
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.059963004990234e-08,
+      "loss": 0.1879,
+      "step": 539
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.344769827770882e-08,
+      "loss": 0.2409,
+      "step": 540
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.662682076050031e-08,
+      "loss": 0.237,
+      "step": 541
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.013722497710817e-08,
+      "loss": 0.2279,
+      "step": 542
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.397912735799371e-08,
+      "loss": 0.215,
+      "step": 543
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.815273327803183e-08,
+      "loss": 0.215,
+      "step": 544
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.2658237049655325e-08,
+      "loss": 0.238,
+      "step": 545
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.7495821916382347e-08,
+      "loss": 0.2373,
+      "step": 546
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.266566004670013e-08,
+      "loss": 0.1975,
+      "step": 547
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.816791252832518e-08,
+      "loss": 0.2247,
+      "step": 548
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.400272936283088e-08,
+      "loss": 0.2064,
+      "step": 549
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.01702494606415e-08,
+      "loss": 0.2157,
+      "step": 550
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.667060063640369e-08,
+      "loss": 0.2034,
+      "step": 551
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.3503899604725424e-08,
+      "loss": 0.2026,
+      "step": 552
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.0670251976275803e-08,
+      "loss": 0.2261,
+      "step": 553
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.169752254270081e-09,
+      "loss": 0.2229,
+      "step": 554
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 6.0024838313144095e-09,
+      "loss": 0.2367,
+      "step": 555
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.1685189866280676e-09,
+      "loss": 0.2111,
+      "step": 556
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.667918883627607e-09,
+      "loss": 0.2385,
+      "step": 557
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.500733567890711e-09,
+      "loss": 0.2067,
+      "step": 558
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 6.670019654875237e-10,
+      "loss": 0.1983,
+      "step": 559
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.6675188168169266e-10,
+      "loss": 0.2235,
+      "step": 560
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0,
+      "loss": 0.2054,
+      "step": 561
+    },
+    {
+      "epoch": 3.0,
+      "step": 561,
+      "total_flos": 7.557759207819182e+17,
+      "train_loss": 0.3695768022813474,
+      "train_runtime": 19005.2972,
+      "train_samples_per_second": 3.782,
+      "train_steps_per_second": 0.03
+    }
+  ],
+  "max_steps": 561,
+  "num_train_epochs": 3,
+  "total_flos": 7.557759207819182e+17,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e49a25caa24acff4db622fa81e7c237fca6f6212c6ae63fb5d7ff56b3a435a8
+size 3707