Upload folder using huggingface_hub

.gitattributes

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

added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "Qwen/Qwen2.5-32B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 5120,
+  "initializer_range": 0.02,
+  "intermediate_size": 27648,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 70,
+  "model_type": "qwen2",
+  "num_attention_heads": 40,
+  "num_hidden_layers": 64,
+  "num_key_value_heads": 8,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.46.1",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.46.1"
+}

latest

@@ -0,0 +1 @@
+global_step381

model-00001-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4aa81886f985ca932799d09b723f4033df00b6986fb4bf6be266c053335dde31
+size 4891730992

model-00002-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2d48d0e6a74a2a042ce807591544964478bf4f4486118ce44828f7caf9199285
+size 4876059352

model-00003-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:353913a21fdabd2e41594b4a3979a81dfe0670a8087b2db27577a47effc45146
+size 4876059384

model-00004-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c8fbba0d6263ea2e911364962109c2e43a05d85cf3759ca70b93659391588a8
+size 4876059416

model-00005-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:81ebdb798d92403f3923250addedec533e7e7c273c0e5bbf6b5bf279c2855554
+size 4876059416

model-00006-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76c061193b77c3e353978d897bf00a05b5a64ab02add549d933ce68823074c4d
+size 4876059416

model-00007-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:746e6d56f9e98fe40aadafdb9c01d3df96f6afe63f6ba50e63f66e03df909da9
+size 4876059416

model-00008-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f9450e588ebf83a2f84eba7c2e75a0d77f42dd68a5a197f7bb335320b36d674
+size 4876059416

model-00009-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a0fcf2ddb9293325222407876e8f4910de64c7d8e95d3092b9ce485a4939d09
+size 4876059416

model-00010-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85206662ffc04a8ebbb6c0f1de21fbd46619597122ffbf9375c016c8817641b3
+size 4876059416

model-00011-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f149485c29637d3eef352e11a67a2e6d80593dd91efb4481c27b555212b8484
+size 4876059416

model-00012-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d43e8d9f470d1bc9fd5b27fbb747b8c0a683fb1a7dafe0dca34770ce413609c0
+size 4876059416

model-00013-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f678ff3838230ea7ddb45eba6114a3ed598174ed696b017aaa41233c27ae35e
+size 4876059416

model-00014-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3d19d34240471677870ca82eea834532f53598fb2effa64fe6faa677d32ee5c
+size 2123397800

model.safetensors.index.json

@@ -0,0 +1,778 @@
+{
+  "metadata": {
+    "total_size": 65527752704
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00014-of-00014.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00014.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00014.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00014.safetensors",
+    "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00014.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00014.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00014.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00014.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00014.safetensors",
+    "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00014.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00014.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00014.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00003-of-00014.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00003-of-00014.safetensors",
+    "model.layers.10.self_attn.k_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.10.self_attn.q_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.10.self_attn.v_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00003-of-00014.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00003-of-00014.safetensors",
+    "model.layers.11.self_attn.k_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.11.self_attn.q_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.11.self_attn.v_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00003-of-00014.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00003-of-00014.safetensors",
+    "model.layers.12.self_attn.k_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.12.self_attn.q_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.12.self_attn.v_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00004-of-00014.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00004-of-00014.safetensors",
+    "model.layers.13.self_attn.k_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.13.self_attn.q_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.13.self_attn.v_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00004-of-00014.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00004-of-00014.safetensors",
+    "model.layers.14.self_attn.k_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.14.self_attn.q_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.14.self_attn.v_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00004-of-00014.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00004-of-00014.safetensors",
+    "model.layers.15.self_attn.k_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.15.self_attn.q_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.15.self_attn.v_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00004-of-00014.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00004-of-00014.safetensors",
+    "model.layers.16.self_attn.k_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.16.self_attn.q_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.16.self_attn.v_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00004-of-00014.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00004-of-00014.safetensors",
+    "model.layers.17.self_attn.k_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.17.self_attn.q_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.17.self_attn.v_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00005-of-00014.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00005-of-00014.safetensors",
+    "model.layers.18.self_attn.k_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.18.self_attn.q_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.18.self_attn.v_proj.bias": "model-00004-of-00014.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00004-of-00014.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00005-of-00014.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00005-of-00014.safetensors",
+    "model.layers.19.self_attn.k_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.19.self_attn.q_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.19.self_attn.v_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00014.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00014.safetensors",
+    "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00014.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.2.self_attn.q_proj.bias": "model-00001-of-00014.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00014.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00005-of-00014.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00005-of-00014.safetensors",
+    "model.layers.20.self_attn.k_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.20.self_attn.q_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.20.self_attn.v_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00005-of-00014.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00005-of-00014.safetensors",
+    "model.layers.21.self_attn.k_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.21.self_attn.q_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.21.self_attn.v_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00005-of-00014.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00005-of-00014.safetensors",
+    "model.layers.22.self_attn.k_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.22.self_attn.q_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.22.self_attn.v_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00006-of-00014.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00006-of-00014.safetensors",
+    "model.layers.23.self_attn.k_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.23.self_attn.q_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.23.self_attn.v_proj.bias": "model-00005-of-00014.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00005-of-00014.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00006-of-00014.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00006-of-00014.safetensors",
+    "model.layers.24.self_attn.k_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.24.self_attn.q_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.24.self_attn.v_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00006-of-00014.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00006-of-00014.safetensors",
+    "model.layers.25.self_attn.k_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.25.self_attn.q_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.25.self_attn.v_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00006-of-00014.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00006-of-00014.safetensors",
+    "model.layers.26.self_attn.k_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.26.self_attn.q_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.26.self_attn.v_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00006-of-00014.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00006-of-00014.safetensors",
+    "model.layers.27.self_attn.k_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.27.self_attn.q_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.27.self_attn.v_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00007-of-00014.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00007-of-00014.safetensors",
+    "model.layers.28.self_attn.k_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.28.self_attn.q_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.28.self_attn.v_proj.bias": "model-00006-of-00014.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00006-of-00014.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00007-of-00014.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00007-of-00014.safetensors",
+    "model.layers.29.self_attn.k_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.29.self_attn.q_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.29.self_attn.v_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00002-of-00014.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00002-of-00014.safetensors",
+    "model.layers.3.self_attn.k_proj.bias": "model-00001-of-00014.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.3.self_attn.q_proj.bias": "model-00001-of-00014.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00014.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00014.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00007-of-00014.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00007-of-00014.safetensors",
+    "model.layers.30.self_attn.k_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.30.self_attn.q_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.30.self_attn.v_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00007-of-00014.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00007-of-00014.safetensors",
+    "model.layers.31.self_attn.k_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.31.self_attn.q_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.31.self_attn.v_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00007-of-00014.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00007-of-00014.safetensors",
+    "model.layers.32.self_attn.k_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.32.self_attn.q_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.32.self_attn.v_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00008-of-00014.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00008-of-00014.safetensors",
+    "model.layers.33.self_attn.k_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.33.self_attn.q_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.33.self_attn.v_proj.bias": "model-00007-of-00014.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00007-of-00014.safetensors",
+    "model.layers.34.input_layernorm.weight": "model-00008-of-00014.safetensors",
+    "model.layers.34.mlp.down_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.34.mlp.gate_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.34.mlp.up_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "model-00008-of-00014.safetensors",
+    "model.layers.34.self_attn.k_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.34.self_attn.k_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.34.self_attn.o_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.34.self_attn.q_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.34.self_attn.q_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.34.self_attn.v_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.34.self_attn.v_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.35.input_layernorm.weight": "model-00008-of-00014.safetensors",
+    "model.layers.35.mlp.down_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.35.mlp.gate_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.35.mlp.up_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "model-00008-of-00014.safetensors",
+    "model.layers.35.self_attn.k_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.35.self_attn.k_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.35.self_attn.o_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.35.self_attn.q_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.35.self_attn.q_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.35.self_attn.v_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.35.self_attn.v_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.36.input_layernorm.weight": "model-00008-of-00014.safetensors",
+    "model.layers.36.mlp.down_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.36.mlp.gate_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.36.mlp.up_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.36.post_attention_layernorm.weight": "model-00008-of-00014.safetensors",
+    "model.layers.36.self_attn.k_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.36.self_attn.k_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.36.self_attn.o_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.36.self_attn.q_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.36.self_attn.q_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.36.self_attn.v_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.36.self_attn.v_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.37.input_layernorm.weight": "model-00008-of-00014.safetensors",
+    "model.layers.37.mlp.down_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.37.mlp.gate_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.37.mlp.up_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.37.post_attention_layernorm.weight": "model-00008-of-00014.safetensors",
+    "model.layers.37.self_attn.k_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.37.self_attn.k_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.37.self_attn.o_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.37.self_attn.q_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.37.self_attn.q_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.37.self_attn.v_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.37.self_attn.v_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.38.input_layernorm.weight": "model-00009-of-00014.safetensors",
+    "model.layers.38.mlp.down_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.38.mlp.gate_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.38.mlp.up_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.38.post_attention_layernorm.weight": "model-00009-of-00014.safetensors",
+    "model.layers.38.self_attn.k_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.38.self_attn.k_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.38.self_attn.o_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.38.self_attn.q_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.38.self_attn.q_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.38.self_attn.v_proj.bias": "model-00008-of-00014.safetensors",
+    "model.layers.38.self_attn.v_proj.weight": "model-00008-of-00014.safetensors",
+    "model.layers.39.input_layernorm.weight": "model-00009-of-00014.safetensors",
+    "model.layers.39.mlp.down_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.39.mlp.gate_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.39.mlp.up_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.39.post_attention_layernorm.weight": "model-00009-of-00014.safetensors",
+    "model.layers.39.self_attn.k_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.39.self_attn.k_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.39.self_attn.o_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.39.self_attn.q_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.39.self_attn.q_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.39.self_attn.v_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.39.self_attn.v_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00002-of-00014.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00002-of-00014.safetensors",
+    "model.layers.4.self_attn.k_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.4.self_attn.q_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.4.self_attn.v_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.40.input_layernorm.weight": "model-00009-of-00014.safetensors",
+    "model.layers.40.mlp.down_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.40.mlp.gate_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.40.mlp.up_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.40.post_attention_layernorm.weight": "model-00009-of-00014.safetensors",
+    "model.layers.40.self_attn.k_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.40.self_attn.k_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.40.self_attn.o_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.40.self_attn.q_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.40.self_attn.q_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.40.self_attn.v_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.40.self_attn.v_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.41.input_layernorm.weight": "model-00009-of-00014.safetensors",
+    "model.layers.41.mlp.down_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.41.mlp.gate_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.41.mlp.up_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.41.post_attention_layernorm.weight": "model-00009-of-00014.safetensors",
+    "model.layers.41.self_attn.k_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.41.self_attn.k_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.41.self_attn.o_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.41.self_attn.q_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.41.self_attn.q_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.41.self_attn.v_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.41.self_attn.v_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.42.input_layernorm.weight": "model-00009-of-00014.safetensors",
+    "model.layers.42.mlp.down_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.42.mlp.gate_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.42.mlp.up_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.42.post_attention_layernorm.weight": "model-00009-of-00014.safetensors",
+    "model.layers.42.self_attn.k_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.42.self_attn.k_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.42.self_attn.o_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.42.self_attn.q_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.42.self_attn.q_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.42.self_attn.v_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.42.self_attn.v_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.43.input_layernorm.weight": "model-00010-of-00014.safetensors",
+    "model.layers.43.mlp.down_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.43.mlp.gate_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.43.mlp.up_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.43.post_attention_layernorm.weight": "model-00010-of-00014.safetensors",
+    "model.layers.43.self_attn.k_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.43.self_attn.k_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.43.self_attn.o_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.43.self_attn.q_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.43.self_attn.q_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.43.self_attn.v_proj.bias": "model-00009-of-00014.safetensors",
+    "model.layers.43.self_attn.v_proj.weight": "model-00009-of-00014.safetensors",
+    "model.layers.44.input_layernorm.weight": "model-00010-of-00014.safetensors",
+    "model.layers.44.mlp.down_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.44.mlp.gate_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.44.mlp.up_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.44.post_attention_layernorm.weight": "model-00010-of-00014.safetensors",
+    "model.layers.44.self_attn.k_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.44.self_attn.k_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.44.self_attn.o_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.44.self_attn.q_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.44.self_attn.q_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.44.self_attn.v_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.44.self_attn.v_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.45.input_layernorm.weight": "model-00010-of-00014.safetensors",
+    "model.layers.45.mlp.down_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.45.mlp.gate_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.45.mlp.up_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.45.post_attention_layernorm.weight": "model-00010-of-00014.safetensors",
+    "model.layers.45.self_attn.k_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.45.self_attn.k_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.45.self_attn.o_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.45.self_attn.q_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.45.self_attn.q_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.45.self_attn.v_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.45.self_attn.v_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.46.input_layernorm.weight": "model-00010-of-00014.safetensors",
+    "model.layers.46.mlp.down_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.46.mlp.gate_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.46.mlp.up_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.46.post_attention_layernorm.weight": "model-00010-of-00014.safetensors",
+    "model.layers.46.self_attn.k_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.46.self_attn.k_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.46.self_attn.o_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.46.self_attn.q_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.46.self_attn.q_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.46.self_attn.v_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.46.self_attn.v_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.47.input_layernorm.weight": "model-00010-of-00014.safetensors",
+    "model.layers.47.mlp.down_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.47.mlp.gate_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.47.mlp.up_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.47.post_attention_layernorm.weight": "model-00010-of-00014.safetensors",
+    "model.layers.47.self_attn.k_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.47.self_attn.k_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.47.self_attn.o_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.47.self_attn.q_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.47.self_attn.q_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.47.self_attn.v_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.47.self_attn.v_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.48.input_layernorm.weight": "model-00011-of-00014.safetensors",
+    "model.layers.48.mlp.down_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.48.mlp.gate_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.48.mlp.up_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.48.post_attention_layernorm.weight": "model-00011-of-00014.safetensors",
+    "model.layers.48.self_attn.k_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.48.self_attn.k_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.48.self_attn.o_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.48.self_attn.q_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.48.self_attn.q_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.48.self_attn.v_proj.bias": "model-00010-of-00014.safetensors",
+    "model.layers.48.self_attn.v_proj.weight": "model-00010-of-00014.safetensors",
+    "model.layers.49.input_layernorm.weight": "model-00011-of-00014.safetensors",
+    "model.layers.49.mlp.down_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.49.mlp.gate_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.49.mlp.up_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.49.post_attention_layernorm.weight": "model-00011-of-00014.safetensors",
+    "model.layers.49.self_attn.k_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.49.self_attn.k_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.49.self_attn.o_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.49.self_attn.q_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.49.self_attn.q_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.49.self_attn.v_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.49.self_attn.v_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00002-of-00014.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00002-of-00014.safetensors",
+    "model.layers.5.self_attn.k_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.5.self_attn.q_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.5.self_attn.v_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.50.input_layernorm.weight": "model-00011-of-00014.safetensors",
+    "model.layers.50.mlp.down_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.50.mlp.gate_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.50.mlp.up_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.50.post_attention_layernorm.weight": "model-00011-of-00014.safetensors",
+    "model.layers.50.self_attn.k_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.50.self_attn.k_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.50.self_attn.o_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.50.self_attn.q_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.50.self_attn.q_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.50.self_attn.v_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.50.self_attn.v_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.51.input_layernorm.weight": "model-00011-of-00014.safetensors",
+    "model.layers.51.mlp.down_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.51.mlp.gate_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.51.mlp.up_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.51.post_attention_layernorm.weight": "model-00011-of-00014.safetensors",
+    "model.layers.51.self_attn.k_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.51.self_attn.k_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.51.self_attn.o_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.51.self_attn.q_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.51.self_attn.q_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.51.self_attn.v_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.51.self_attn.v_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.52.input_layernorm.weight": "model-00011-of-00014.safetensors",
+    "model.layers.52.mlp.down_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.52.mlp.gate_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.52.mlp.up_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.52.post_attention_layernorm.weight": "model-00011-of-00014.safetensors",
+    "model.layers.52.self_attn.k_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.52.self_attn.k_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.52.self_attn.o_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.52.self_attn.q_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.52.self_attn.q_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.52.self_attn.v_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.52.self_attn.v_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.53.input_layernorm.weight": "model-00012-of-00014.safetensors",
+    "model.layers.53.mlp.down_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.53.mlp.gate_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.53.mlp.up_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.53.post_attention_layernorm.weight": "model-00012-of-00014.safetensors",
+    "model.layers.53.self_attn.k_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.53.self_attn.k_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.53.self_attn.o_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.53.self_attn.q_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.53.self_attn.q_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.53.self_attn.v_proj.bias": "model-00011-of-00014.safetensors",
+    "model.layers.53.self_attn.v_proj.weight": "model-00011-of-00014.safetensors",
+    "model.layers.54.input_layernorm.weight": "model-00012-of-00014.safetensors",
+    "model.layers.54.mlp.down_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.54.mlp.gate_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.54.mlp.up_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.54.post_attention_layernorm.weight": "model-00012-of-00014.safetensors",
+    "model.layers.54.self_attn.k_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.54.self_attn.k_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.54.self_attn.o_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.54.self_attn.q_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.54.self_attn.q_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.54.self_attn.v_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.54.self_attn.v_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.55.input_layernorm.weight": "model-00012-of-00014.safetensors",
+    "model.layers.55.mlp.down_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.55.mlp.gate_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.55.mlp.up_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.55.post_attention_layernorm.weight": "model-00012-of-00014.safetensors",
+    "model.layers.55.self_attn.k_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.55.self_attn.k_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.55.self_attn.o_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.55.self_attn.q_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.55.self_attn.q_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.55.self_attn.v_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.55.self_attn.v_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.56.input_layernorm.weight": "model-00012-of-00014.safetensors",
+    "model.layers.56.mlp.down_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.56.mlp.gate_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.56.mlp.up_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.56.post_attention_layernorm.weight": "model-00012-of-00014.safetensors",
+    "model.layers.56.self_attn.k_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.56.self_attn.k_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.56.self_attn.o_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.56.self_attn.q_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.56.self_attn.q_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.56.self_attn.v_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.56.self_attn.v_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.57.input_layernorm.weight": "model-00012-of-00014.safetensors",
+    "model.layers.57.mlp.down_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.57.mlp.gate_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.57.mlp.up_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.57.post_attention_layernorm.weight": "model-00012-of-00014.safetensors",
+    "model.layers.57.self_attn.k_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.57.self_attn.k_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.57.self_attn.o_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.57.self_attn.q_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.57.self_attn.q_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.57.self_attn.v_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.57.self_attn.v_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.58.input_layernorm.weight": "model-00013-of-00014.safetensors",
+    "model.layers.58.mlp.down_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.58.mlp.gate_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.58.mlp.up_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.58.post_attention_layernorm.weight": "model-00013-of-00014.safetensors",
+    "model.layers.58.self_attn.k_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.58.self_attn.k_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.58.self_attn.o_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.58.self_attn.q_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.58.self_attn.q_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.58.self_attn.v_proj.bias": "model-00012-of-00014.safetensors",
+    "model.layers.58.self_attn.v_proj.weight": "model-00012-of-00014.safetensors",
+    "model.layers.59.input_layernorm.weight": "model-00013-of-00014.safetensors",
+    "model.layers.59.mlp.down_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.59.mlp.gate_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.59.mlp.up_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.59.post_attention_layernorm.weight": "model-00013-of-00014.safetensors",
+    "model.layers.59.self_attn.k_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.59.self_attn.k_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.59.self_attn.o_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.59.self_attn.q_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.59.self_attn.q_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.59.self_attn.v_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.59.self_attn.v_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00002-of-00014.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00002-of-00014.safetensors",
+    "model.layers.6.self_attn.k_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.6.self_attn.q_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.6.self_attn.v_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.60.input_layernorm.weight": "model-00013-of-00014.safetensors",
+    "model.layers.60.mlp.down_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.60.mlp.gate_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.60.mlp.up_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.60.post_attention_layernorm.weight": "model-00013-of-00014.safetensors",
+    "model.layers.60.self_attn.k_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.60.self_attn.k_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.60.self_attn.o_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.60.self_attn.q_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.60.self_attn.q_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.60.self_attn.v_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.60.self_attn.v_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.61.input_layernorm.weight": "model-00013-of-00014.safetensors",
+    "model.layers.61.mlp.down_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.61.mlp.gate_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.61.mlp.up_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.61.post_attention_layernorm.weight": "model-00013-of-00014.safetensors",
+    "model.layers.61.self_attn.k_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.61.self_attn.k_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.61.self_attn.o_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.61.self_attn.q_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.61.self_attn.q_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.61.self_attn.v_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.61.self_attn.v_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.62.input_layernorm.weight": "model-00013-of-00014.safetensors",
+    "model.layers.62.mlp.down_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.62.mlp.gate_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.62.mlp.up_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.62.post_attention_layernorm.weight": "model-00013-of-00014.safetensors",
+    "model.layers.62.self_attn.k_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.62.self_attn.k_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.62.self_attn.o_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.62.self_attn.q_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.62.self_attn.q_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.62.self_attn.v_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.62.self_attn.v_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.63.input_layernorm.weight": "model-00014-of-00014.safetensors",
+    "model.layers.63.mlp.down_proj.weight": "model-00014-of-00014.safetensors",
+    "model.layers.63.mlp.gate_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.63.mlp.up_proj.weight": "model-00014-of-00014.safetensors",
+    "model.layers.63.post_attention_layernorm.weight": "model-00014-of-00014.safetensors",
+    "model.layers.63.self_attn.k_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.63.self_attn.k_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.63.self_attn.o_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.63.self_attn.q_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.63.self_attn.q_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.63.self_attn.v_proj.bias": "model-00013-of-00014.safetensors",
+    "model.layers.63.self_attn.v_proj.weight": "model-00013-of-00014.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00002-of-00014.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00002-of-00014.safetensors",
+    "model.layers.7.self_attn.k_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.7.self_attn.q_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.7.self_attn.v_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00003-of-00014.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00003-of-00014.safetensors",
+    "model.layers.8.self_attn.k_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.8.self_attn.q_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.8.self_attn.v_proj.bias": "model-00002-of-00014.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00002-of-00014.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00003-of-00014.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00003-of-00014.safetensors",
+    "model.layers.9.self_attn.k_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.9.self_attn.q_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00003-of-00014.safetensors",
+    "model.layers.9.self_attn.v_proj.bias": "model-00003-of-00014.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00003-of-00014.safetensors",
+    "model.norm.weight": "model-00014-of-00014.safetensors"
+  }
+}

rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08282b46825aa78d10fe10e3fea89555c5b5a691b261a3ddfd58fcb58370edff
+size 15984

rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbab71d98a3a9a92df82a6bba463947327c3a1bcf35cd9f4f46114641fc42dd9
+size 15984

rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:caac82d57d878d30219a4f9ec289a97ff90c53afc160b968f251b3fd3454b8d8
+size 15984

rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:19762d2d370222b01817da11bbaa6665d542293373186d66f754e7246bb861ed
+size 15984

rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:00c7508b346a7d3c5c23392845f1d013331114ade778794b76e919cb3ed5d33e
+size 15984

rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b89de7d14dd20a191f56b74c816ef8b7fe5c171e31efbeadbf321c4539ed68c3
+size 15984

rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c71152053553e6e22d670fbc4fd7550bf8a046b54cad7b71869787986a6a42c
+size 15984

rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b67db12a26a26ffe03d9afc84a43857eb2e5b2fec2dd189653b415f74208190
+size 15984

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88360606f3db9d57bd14ab89d27da403ae3cb4d6eeb3c471fd1c57b4815a9868
+size 1064

special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

tokenizer_config.json

@@ -0,0 +1,208 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "model_max_length": 16384,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

trainer_state.json

@@ -0,0 +1,2700 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.990202482037884,
+  "eval_steps": 500,
+  "global_step": 381,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.007838014369693011,
+      "grad_norm": 3.012711763381958,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 0.5662,
+      "step": 1
+    },
+    {
+      "epoch": 0.015676028739386023,
+      "grad_norm": 3.1244778633117676,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 0.5951,
+      "step": 2
+    },
+    {
+      "epoch": 0.023514043109079032,
+      "grad_norm": 3.0221784114837646,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.5891,
+      "step": 3
+    },
+    {
+      "epoch": 0.031352057478772045,
+      "grad_norm": 2.6240241527557373,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.5812,
+      "step": 4
+    },
+    {
+      "epoch": 0.039190071848465055,
+      "grad_norm": 1.920555591583252,
+      "learning_rate": 6.41025641025641e-06,
+      "loss": 0.54,
+      "step": 5
+    },
+    {
+      "epoch": 0.047028086218158065,
+      "grad_norm": 1.4635577201843262,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.5043,
+      "step": 6
+    },
+    {
+      "epoch": 0.054866100587851074,
+      "grad_norm": 1.418139100074768,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 0.491,
+      "step": 7
+    },
+    {
+      "epoch": 0.06270411495754409,
+      "grad_norm": 1.3632681369781494,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 0.4744,
+      "step": 8
+    },
+    {
+      "epoch": 0.0705421293272371,
+      "grad_norm": 1.7710790634155273,
+      "learning_rate": 1.153846153846154e-05,
+      "loss": 0.4128,
+      "step": 9
+    },
+    {
+      "epoch": 0.07838014369693011,
+      "grad_norm": 1.3074016571044922,
+      "learning_rate": 1.282051282051282e-05,
+      "loss": 0.3723,
+      "step": 10
+    },
+    {
+      "epoch": 0.08621815806662313,
+      "grad_norm": 0.8031821250915527,
+      "learning_rate": 1.4102564102564104e-05,
+      "loss": 0.3585,
+      "step": 11
+    },
+    {
+      "epoch": 0.09405617243631613,
+      "grad_norm": 1.116591453552246,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.3427,
+      "step": 12
+    },
+    {
+      "epoch": 0.10189418680600915,
+      "grad_norm": 0.970167875289917,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.3333,
+      "step": 13
+    },
+    {
+      "epoch": 0.10973220117570215,
+      "grad_norm": 0.8765995502471924,
+      "learning_rate": 1.794871794871795e-05,
+      "loss": 0.3051,
+      "step": 14
+    },
+    {
+      "epoch": 0.11757021554539517,
+      "grad_norm": 0.7228055000305176,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 0.2992,
+      "step": 15
+    },
+    {
+      "epoch": 0.12540822991508818,
+      "grad_norm": 0.6161549687385559,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 0.2884,
+      "step": 16
+    },
+    {
+      "epoch": 0.13324624428478118,
+      "grad_norm": 0.5537469983100891,
+      "learning_rate": 2.1794871794871795e-05,
+      "loss": 0.2829,
+      "step": 17
+    },
+    {
+      "epoch": 0.1410842586544742,
+      "grad_norm": 0.5031456351280212,
+      "learning_rate": 2.307692307692308e-05,
+      "loss": 0.2651,
+      "step": 18
+    },
+    {
+      "epoch": 0.14892227302416722,
+      "grad_norm": 0.49946239590644836,
+      "learning_rate": 2.435897435897436e-05,
+      "loss": 0.2594,
+      "step": 19
+    },
+    {
+      "epoch": 0.15676028739386022,
+      "grad_norm": 0.434522420167923,
+      "learning_rate": 2.564102564102564e-05,
+      "loss": 0.2491,
+      "step": 20
+    },
+    {
+      "epoch": 0.16459830176355322,
+      "grad_norm": 0.5493834614753723,
+      "learning_rate": 2.6923076923076923e-05,
+      "loss": 0.2474,
+      "step": 21
+    },
+    {
+      "epoch": 0.17243631613324625,
+      "grad_norm": 0.40903612971305847,
+      "learning_rate": 2.8205128205128207e-05,
+      "loss": 0.2534,
+      "step": 22
+    },
+    {
+      "epoch": 0.18027433050293926,
+      "grad_norm": 0.46355342864990234,
+      "learning_rate": 2.948717948717949e-05,
+      "loss": 0.2547,
+      "step": 23
+    },
+    {
+      "epoch": 0.18811234487263226,
+      "grad_norm": 0.38935673236846924,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 0.238,
+      "step": 24
+    },
+    {
+      "epoch": 0.1959503592423253,
+      "grad_norm": 0.43248051404953003,
+      "learning_rate": 3.205128205128206e-05,
+      "loss": 0.243,
+      "step": 25
+    },
+    {
+      "epoch": 0.2037883736120183,
+      "grad_norm": 0.3940970301628113,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.2365,
+      "step": 26
+    },
+    {
+      "epoch": 0.2116263879817113,
+      "grad_norm": 0.39611291885375977,
+      "learning_rate": 3.461538461538462e-05,
+      "loss": 0.2299,
+      "step": 27
+    },
+    {
+      "epoch": 0.2194644023514043,
+      "grad_norm": 0.36743637919425964,
+      "learning_rate": 3.58974358974359e-05,
+      "loss": 0.235,
+      "step": 28
+    },
+    {
+      "epoch": 0.22730241672109733,
+      "grad_norm": 0.35339489579200745,
+      "learning_rate": 3.717948717948718e-05,
+      "loss": 0.224,
+      "step": 29
+    },
+    {
+      "epoch": 0.23514043109079033,
+      "grad_norm": 0.33060234785079956,
+      "learning_rate": 3.846153846153846e-05,
+      "loss": 0.2276,
+      "step": 30
+    },
+    {
+      "epoch": 0.24297844546048333,
+      "grad_norm": 0.3493947386741638,
+      "learning_rate": 3.974358974358974e-05,
+      "loss": 0.228,
+      "step": 31
+    },
+    {
+      "epoch": 0.25081645983017636,
+      "grad_norm": 0.34183913469314575,
+      "learning_rate": 4.1025641025641023e-05,
+      "loss": 0.2245,
+      "step": 32
+    },
+    {
+      "epoch": 0.25865447419986937,
+      "grad_norm": 0.3201765716075897,
+      "learning_rate": 4.230769230769231e-05,
+      "loss": 0.2303,
+      "step": 33
+    },
+    {
+      "epoch": 0.26649248856956237,
+      "grad_norm": 0.3946562707424164,
+      "learning_rate": 4.358974358974359e-05,
+      "loss": 0.2266,
+      "step": 34
+    },
+    {
+      "epoch": 0.27433050293925537,
+      "grad_norm": 0.36500218510627747,
+      "learning_rate": 4.4871794871794874e-05,
+      "loss": 0.2103,
+      "step": 35
+    },
+    {
+      "epoch": 0.2821685173089484,
+      "grad_norm": 0.3446497321128845,
+      "learning_rate": 4.615384615384616e-05,
+      "loss": 0.2309,
+      "step": 36
+    },
+    {
+      "epoch": 0.29000653167864143,
+      "grad_norm": 0.3353784382343292,
+      "learning_rate": 4.7435897435897435e-05,
+      "loss": 0.2294,
+      "step": 37
+    },
+    {
+      "epoch": 0.29784454604833444,
+      "grad_norm": 0.3502955138683319,
+      "learning_rate": 4.871794871794872e-05,
+      "loss": 0.2279,
+      "step": 38
+    },
+    {
+      "epoch": 0.30568256041802744,
+      "grad_norm": 0.3561626076698303,
+      "learning_rate": 5e-05,
+      "loss": 0.2289,
+      "step": 39
+    },
+    {
+      "epoch": 0.31352057478772044,
+      "grad_norm": 0.3842853009700775,
+      "learning_rate": 4.999894523795781e-05,
+      "loss": 0.2257,
+      "step": 40
+    },
+    {
+      "epoch": 0.32135858915741344,
+      "grad_norm": 0.4691725969314575,
+      "learning_rate": 4.999578104083307e-05,
+      "loss": 0.2227,
+      "step": 41
+    },
+    {
+      "epoch": 0.32919660352710645,
+      "grad_norm": 0.4886467754840851,
+      "learning_rate": 4.999050767562379e-05,
+      "loss": 0.2187,
+      "step": 42
+    },
+    {
+      "epoch": 0.3370346178967995,
+      "grad_norm": 0.38050761818885803,
+      "learning_rate": 4.998312558730159e-05,
+      "loss": 0.2301,
+      "step": 43
+    },
+    {
+      "epoch": 0.3448726322664925,
+      "grad_norm": 0.33100268244743347,
+      "learning_rate": 4.997363539877422e-05,
+      "loss": 0.2289,
+      "step": 44
+    },
+    {
+      "epoch": 0.3527106466361855,
+      "grad_norm": 0.36914992332458496,
+      "learning_rate": 4.996203791083291e-05,
+      "loss": 0.2301,
+      "step": 45
+    },
+    {
+      "epoch": 0.3605486610058785,
+      "grad_norm": 0.32373785972595215,
+      "learning_rate": 4.994833410208487e-05,
+      "loss": 0.2212,
+      "step": 46
+    },
+    {
+      "epoch": 0.3683866753755715,
+      "grad_norm": 0.3918166160583496,
+      "learning_rate": 4.993252512887069e-05,
+      "loss": 0.229,
+      "step": 47
+    },
+    {
+      "epoch": 0.3762246897452645,
+      "grad_norm": 0.3503532111644745,
+      "learning_rate": 4.991461232516675e-05,
+      "loss": 0.2323,
+      "step": 48
+    },
+    {
+      "epoch": 0.3840627041149575,
+      "grad_norm": 0.3073519766330719,
+      "learning_rate": 4.9894597202472696e-05,
+      "loss": 0.2248,
+      "step": 49
+    },
+    {
+      "epoch": 0.3919007184846506,
+      "grad_norm": 0.30389830470085144,
+      "learning_rate": 4.9872481449683844e-05,
+      "loss": 0.234,
+      "step": 50
+    },
+    {
+      "epoch": 0.3997387328543436,
+      "grad_norm": 0.3088521957397461,
+      "learning_rate": 4.984826693294874e-05,
+      "loss": 0.232,
+      "step": 51
+    },
+    {
+      "epoch": 0.4075767472240366,
+      "grad_norm": 0.2747111916542053,
+      "learning_rate": 4.982195569551162e-05,
+      "loss": 0.238,
+      "step": 52
+    },
+    {
+      "epoch": 0.4154147615937296,
+      "grad_norm": 0.33293288946151733,
+      "learning_rate": 4.979354995754006e-05,
+      "loss": 0.2407,
+      "step": 53
+    },
+    {
+      "epoch": 0.4232527759634226,
+      "grad_norm": 0.3552456796169281,
+      "learning_rate": 4.976305211593758e-05,
+      "loss": 0.2422,
+      "step": 54
+    },
+    {
+      "epoch": 0.4310907903331156,
+      "grad_norm": 0.29555025696754456,
+      "learning_rate": 4.9730464744141445e-05,
+      "loss": 0.2311,
+      "step": 55
+    },
+    {
+      "epoch": 0.4389288047028086,
+      "grad_norm": 0.314091295003891,
+      "learning_rate": 4.969579059190549e-05,
+      "loss": 0.2329,
+      "step": 56
+    },
+    {
+      "epoch": 0.44676681907250165,
+      "grad_norm": 0.30469128489494324,
+      "learning_rate": 4.965903258506806e-05,
+      "loss": 0.2278,
+      "step": 57
+    },
+    {
+      "epoch": 0.45460483344219466,
+      "grad_norm": 0.32766035199165344,
+      "learning_rate": 4.962019382530521e-05,
+      "loss": 0.2341,
+      "step": 58
+    },
+    {
+      "epoch": 0.46244284781188766,
+      "grad_norm": 0.29133275151252747,
+      "learning_rate": 4.957927758986888e-05,
+      "loss": 0.2274,
+      "step": 59
+    },
+    {
+      "epoch": 0.47028086218158066,
+      "grad_norm": 0.36449989676475525,
+      "learning_rate": 4.953628733131045e-05,
+      "loss": 0.2332,
+      "step": 60
+    },
+    {
+      "epoch": 0.47811887655127366,
+      "grad_norm": 0.2951260805130005,
+      "learning_rate": 4.949122667718935e-05,
+      "loss": 0.2408,
+      "step": 61
+    },
+    {
+      "epoch": 0.48595689092096667,
+      "grad_norm": 0.35234174132347107,
+      "learning_rate": 4.944409942976699e-05,
+      "loss": 0.2372,
+      "step": 62
+    },
+    {
+      "epoch": 0.4937949052906597,
+      "grad_norm": 0.33611589670181274,
+      "learning_rate": 4.9394909565685894e-05,
+      "loss": 0.2377,
+      "step": 63
+    },
+    {
+      "epoch": 0.5016329196603527,
+      "grad_norm": 0.32448795437812805,
+      "learning_rate": 4.93436612356342e-05,
+      "loss": 0.2374,
+      "step": 64
+    },
+    {
+      "epoch": 0.5094709340300457,
+      "grad_norm": 0.3616539537906647,
+      "learning_rate": 4.929035876399535e-05,
+      "loss": 0.2347,
+      "step": 65
+    },
+    {
+      "epoch": 0.5173089483997387,
+      "grad_norm": 0.3518874943256378,
+      "learning_rate": 4.923500664848326e-05,
+      "loss": 0.2368,
+      "step": 66
+    },
+    {
+      "epoch": 0.5251469627694317,
+      "grad_norm": 0.28374969959259033,
+      "learning_rate": 4.917760955976277e-05,
+      "loss": 0.2278,
+      "step": 67
+    },
+    {
+      "epoch": 0.5329849771391247,
+      "grad_norm": 0.3445357382297516,
+      "learning_rate": 4.9118172341055516e-05,
+      "loss": 0.2251,
+      "step": 68
+    },
+    {
+      "epoch": 0.5408229915088177,
+      "grad_norm": 0.31107786297798157,
+      "learning_rate": 4.905670000773126e-05,
+      "loss": 0.2288,
+      "step": 69
+    },
+    {
+      "epoch": 0.5486610058785107,
+      "grad_norm": 0.29442402720451355,
+      "learning_rate": 4.899319774688473e-05,
+      "loss": 0.2373,
+      "step": 70
+    },
+    {
+      "epoch": 0.5564990202482037,
+      "grad_norm": 0.33253952860832214,
+      "learning_rate": 4.892767091689786e-05,
+      "loss": 0.2363,
+      "step": 71
+    },
+    {
+      "epoch": 0.5643370346178967,
+      "grad_norm": 0.28732171654701233,
+      "learning_rate": 4.886012504698769e-05,
+      "loss": 0.2336,
+      "step": 72
+    },
+    {
+      "epoch": 0.5721750489875899,
+      "grad_norm": 0.3209660053253174,
+      "learning_rate": 4.87905658367398e-05,
+      "loss": 0.2444,
+      "step": 73
+    },
+    {
+      "epoch": 0.5800130633572829,
+      "grad_norm": 0.3354484438896179,
+      "learning_rate": 4.871899915562736e-05,
+      "loss": 0.2286,
+      "step": 74
+    },
+    {
+      "epoch": 0.5878510777269759,
+      "grad_norm": 0.34029415249824524,
+      "learning_rate": 4.864543104251587e-05,
+      "loss": 0.2375,
+      "step": 75
+    },
+    {
+      "epoch": 0.5956890920966689,
+      "grad_norm": 0.305865615606308,
+      "learning_rate": 4.856986770515358e-05,
+      "loss": 0.2397,
+      "step": 76
+    },
+    {
+      "epoch": 0.6035271064663619,
+      "grad_norm": 0.31168901920318604,
+      "learning_rate": 4.849231551964771e-05,
+      "loss": 0.2407,
+      "step": 77
+    },
+    {
+      "epoch": 0.6113651208360549,
+      "grad_norm": 0.3203716576099396,
+      "learning_rate": 4.841278102992637e-05,
+      "loss": 0.2305,
+      "step": 78
+    },
+    {
+      "epoch": 0.6192031352057479,
+      "grad_norm": 0.301705539226532,
+      "learning_rate": 4.833127094718643e-05,
+      "loss": 0.2228,
+      "step": 79
+    },
+    {
+      "epoch": 0.6270411495754409,
+      "grad_norm": 0.31093651056289673,
+      "learning_rate": 4.82477921493272e-05,
+      "loss": 0.2282,
+      "step": 80
+    },
+    {
+      "epoch": 0.6348791639451339,
+      "grad_norm": 0.30084165930747986,
+      "learning_rate": 4.8162351680370044e-05,
+      "loss": 0.2252,
+      "step": 81
+    },
+    {
+      "epoch": 0.6427171783148269,
+      "grad_norm": 0.2620241641998291,
+      "learning_rate": 4.8074956749864075e-05,
+      "loss": 0.2401,
+      "step": 82
+    },
+    {
+      "epoch": 0.6505551926845199,
+      "grad_norm": 0.2853231728076935,
+      "learning_rate": 4.79856147322777e-05,
+      "loss": 0.2346,
+      "step": 83
+    },
+    {
+      "epoch": 0.6583932070542129,
+      "grad_norm": 0.28441861271858215,
+      "learning_rate": 4.789433316637644e-05,
+      "loss": 0.2393,
+      "step": 84
+    },
+    {
+      "epoch": 0.6662312214239059,
+      "grad_norm": 0.2583891749382019,
+      "learning_rate": 4.7801119754586766e-05,
+      "loss": 0.2328,
+      "step": 85
+    },
+    {
+      "epoch": 0.674069235793599,
+      "grad_norm": 0.3212043344974518,
+      "learning_rate": 4.7705982362346164e-05,
+      "loss": 0.2336,
+      "step": 86
+    },
+    {
+      "epoch": 0.681907250163292,
+      "grad_norm": 0.29511553049087524,
+      "learning_rate": 4.760892901743944e-05,
+      "loss": 0.223,
+      "step": 87
+    },
+    {
+      "epoch": 0.689745264532985,
+      "grad_norm": 0.2569979429244995,
+      "learning_rate": 4.750996790932134e-05,
+      "loss": 0.227,
+      "step": 88
+    },
+    {
+      "epoch": 0.697583278902678,
+      "grad_norm": 0.3018832504749298,
+      "learning_rate": 4.7409107388425504e-05,
+      "loss": 0.2276,
+      "step": 89
+    },
+    {
+      "epoch": 0.705421293272371,
+      "grad_norm": 0.3047080636024475,
+      "learning_rate": 4.730635596545985e-05,
+      "loss": 0.2246,
+      "step": 90
+    },
+    {
+      "epoch": 0.713259307642064,
+      "grad_norm": 0.2749479413032532,
+      "learning_rate": 4.7201722310688445e-05,
+      "loss": 0.2259,
+      "step": 91
+    },
+    {
+      "epoch": 0.721097322011757,
+      "grad_norm": 0.27786287665367126,
+      "learning_rate": 4.709521525319986e-05,
+      "loss": 0.2294,
+      "step": 92
+    },
+    {
+      "epoch": 0.72893533638145,
+      "grad_norm": 0.26435744762420654,
+      "learning_rate": 4.698684378016222e-05,
+      "loss": 0.2241,
+      "step": 93
+    },
+    {
+      "epoch": 0.736773350751143,
+      "grad_norm": 0.26848137378692627,
+      "learning_rate": 4.6876617036064844e-05,
+      "loss": 0.2278,
+      "step": 94
+    },
+    {
+      "epoch": 0.744611365120836,
+      "grad_norm": 0.2701055407524109,
+      "learning_rate": 4.676454432194656e-05,
+      "loss": 0.2369,
+      "step": 95
+    },
+    {
+      "epoch": 0.752449379490529,
+      "grad_norm": 0.2705919146537781,
+      "learning_rate": 4.665063509461097e-05,
+      "loss": 0.2262,
+      "step": 96
+    },
+    {
+      "epoch": 0.760287393860222,
+      "grad_norm": 0.2773014307022095,
+      "learning_rate": 4.6534898965828405e-05,
+      "loss": 0.2355,
+      "step": 97
+    },
+    {
+      "epoch": 0.768125408229915,
+      "grad_norm": 0.23928780853748322,
+      "learning_rate": 4.6417345701524915e-05,
+      "loss": 0.2273,
+      "step": 98
+    },
+    {
+      "epoch": 0.7759634225996082,
+      "grad_norm": 0.26342830061912537,
+      "learning_rate": 4.629798522095818e-05,
+      "loss": 0.2293,
+      "step": 99
+    },
+    {
+      "epoch": 0.7838014369693012,
+      "grad_norm": 0.2792154550552368,
+      "learning_rate": 4.617682759588055e-05,
+      "loss": 0.2331,
+      "step": 100
+    },
+    {
+      "epoch": 0.7916394513389942,
+      "grad_norm": 0.2701910436153412,
+      "learning_rate": 4.6053883049689145e-05,
+      "loss": 0.2278,
+      "step": 101
+    },
+    {
+      "epoch": 0.7994774657086872,
+      "grad_norm": 0.2860325276851654,
+      "learning_rate": 4.592916195656322e-05,
+      "loss": 0.2342,
+      "step": 102
+    },
+    {
+      "epoch": 0.8073154800783802,
+      "grad_norm": 0.3034144639968872,
+      "learning_rate": 4.580267484058876e-05,
+      "loss": 0.2351,
+      "step": 103
+    },
+    {
+      "epoch": 0.8151534944480732,
+      "grad_norm": 0.28156647086143494,
+      "learning_rate": 4.5674432374870455e-05,
+      "loss": 0.2322,
+      "step": 104
+    },
+    {
+      "epoch": 0.8229915088177662,
+      "grad_norm": 0.2784689962863922,
+      "learning_rate": 4.554444538063113e-05,
+      "loss": 0.2361,
+      "step": 105
+    },
+    {
+      "epoch": 0.8308295231874592,
+      "grad_norm": 0.2945846617221832,
+      "learning_rate": 4.5412724826298576e-05,
+      "loss": 0.2302,
+      "step": 106
+    },
+    {
+      "epoch": 0.8386675375571522,
+      "grad_norm": 0.25669312477111816,
+      "learning_rate": 4.5279281826580056e-05,
+      "loss": 0.2327,
+      "step": 107
+    },
+    {
+      "epoch": 0.8465055519268452,
+      "grad_norm": 0.2950827181339264,
+      "learning_rate": 4.514412764152446e-05,
+      "loss": 0.2332,
+      "step": 108
+    },
+    {
+      "epoch": 0.8543435662965382,
+      "grad_norm": 0.2892044186592102,
+      "learning_rate": 4.5007273675572104e-05,
+      "loss": 0.2286,
+      "step": 109
+    },
+    {
+      "epoch": 0.8621815806662312,
+      "grad_norm": 0.27470967173576355,
+      "learning_rate": 4.48687314765925e-05,
+      "loss": 0.2364,
+      "step": 110
+    },
+    {
+      "epoch": 0.8700195950359242,
+      "grad_norm": 0.27829453349113464,
+      "learning_rate": 4.4728512734909844e-05,
+      "loss": 0.2304,
+      "step": 111
+    },
+    {
+      "epoch": 0.8778576094056172,
+      "grad_norm": 0.29203999042510986,
+      "learning_rate": 4.4586629282316654e-05,
+      "loss": 0.2363,
+      "step": 112
+    },
+    {
+      "epoch": 0.8856956237753103,
+      "grad_norm": 0.24676302075386047,
+      "learning_rate": 4.444309309107535e-05,
+      "loss": 0.2215,
+      "step": 113
+    },
+    {
+      "epoch": 0.8935336381450033,
+      "grad_norm": 0.27994513511657715,
+      "learning_rate": 4.4297916272908024e-05,
+      "loss": 0.232,
+      "step": 114
+    },
+    {
+      "epoch": 0.9013716525146963,
+      "grad_norm": 0.2659918963909149,
+      "learning_rate": 4.415111107797445e-05,
+      "loss": 0.2323,
+      "step": 115
+    },
+    {
+      "epoch": 0.9092096668843893,
+      "grad_norm": 0.2813761234283447,
+      "learning_rate": 4.400268989383841e-05,
+      "loss": 0.2294,
+      "step": 116
+    },
+    {
+      "epoch": 0.9170476812540823,
+      "grad_norm": 0.27070480585098267,
+      "learning_rate": 4.385266524442241e-05,
+      "loss": 0.2254,
+      "step": 117
+    },
+    {
+      "epoch": 0.9248856956237753,
+      "grad_norm": 0.2846162021160126,
+      "learning_rate": 4.370104978895089e-05,
+      "loss": 0.2324,
+      "step": 118
+    },
+    {
+      "epoch": 0.9327237099934683,
+      "grad_norm": 0.257036954164505,
+      "learning_rate": 4.3547856320882044e-05,
+      "loss": 0.23,
+      "step": 119
+    },
+    {
+      "epoch": 0.9405617243631613,
+      "grad_norm": 0.28918588161468506,
+      "learning_rate": 4.3393097766828293e-05,
+      "loss": 0.2234,
+      "step": 120
+    },
+    {
+      "epoch": 0.9483997387328543,
+      "grad_norm": 0.23960325121879578,
+      "learning_rate": 4.3236787185465525e-05,
+      "loss": 0.2222,
+      "step": 121
+    },
+    {
+      "epoch": 0.9562377531025473,
+      "grad_norm": 0.26537278294563293,
+      "learning_rate": 4.307893776643117e-05,
+      "loss": 0.2285,
+      "step": 122
+    },
+    {
+      "epoch": 0.9640757674722403,
+      "grad_norm": 0.23799960315227509,
+      "learning_rate": 4.2919562829211283e-05,
+      "loss": 0.2214,
+      "step": 123
+    },
+    {
+      "epoch": 0.9719137818419333,
+      "grad_norm": 0.2599596083164215,
+      "learning_rate": 4.27586758220166e-05,
+      "loss": 0.2246,
+      "step": 124
+    },
+    {
+      "epoch": 0.9797517962116263,
+      "grad_norm": 0.25664016604423523,
+      "learning_rate": 4.259629032064779e-05,
+      "loss": 0.2223,
+      "step": 125
+    },
+    {
+      "epoch": 0.9875898105813194,
+      "grad_norm": 0.24727745354175568,
+      "learning_rate": 4.2432420027349886e-05,
+      "loss": 0.2225,
+      "step": 126
+    },
+    {
+      "epoch": 0.9954278249510125,
+      "grad_norm": 0.2626395523548126,
+      "learning_rate": 4.226707876965611e-05,
+      "loss": 0.2333,
+      "step": 127
+    },
+    {
+      "epoch": 1.0052253429131286,
+      "grad_norm": 0.4331790506839752,
+      "learning_rate": 4.210028049922108e-05,
+      "loss": 0.3128,
+      "step": 128
+    },
+    {
+      "epoch": 1.0130633572828216,
+      "grad_norm": 0.31224802136421204,
+      "learning_rate": 4.193203929064353e-05,
+      "loss": 0.1265,
+      "step": 129
+    },
+    {
+      "epoch": 1.0209013716525146,
+      "grad_norm": 0.34126603603363037,
+      "learning_rate": 4.176236934027873e-05,
+      "loss": 0.1188,
+      "step": 130
+    },
+    {
+      "epoch": 1.0287393860222076,
+      "grad_norm": 0.3540596067905426,
+      "learning_rate": 4.159128496504053e-05,
+      "loss": 0.1149,
+      "step": 131
+    },
+    {
+      "epoch": 1.0365774003919008,
+      "grad_norm": 0.2591104209423065,
+      "learning_rate": 4.141880060119336e-05,
+      "loss": 0.1109,
+      "step": 132
+    },
+    {
+      "epoch": 1.0444154147615938,
+      "grad_norm": 0.33954110741615295,
+      "learning_rate": 4.1244930803134e-05,
+      "loss": 0.1307,
+      "step": 133
+    },
+    {
+      "epoch": 1.0522534291312868,
+      "grad_norm": 0.2748555839061737,
+      "learning_rate": 4.1069690242163484e-05,
+      "loss": 0.1246,
+      "step": 134
+    },
+    {
+      "epoch": 1.0600914435009798,
+      "grad_norm": 0.29179254174232483,
+      "learning_rate": 4.089309370524921e-05,
+      "loss": 0.1144,
+      "step": 135
+    },
+    {
+      "epoch": 1.0679294578706728,
+      "grad_norm": 0.3250279724597931,
+      "learning_rate": 4.071515609377705e-05,
+      "loss": 0.1363,
+      "step": 136
+    },
+    {
+      "epoch": 1.0757674722403658,
+      "grad_norm": 0.24014687538146973,
+      "learning_rate": 4.053589242229412e-05,
+      "loss": 0.1044,
+      "step": 137
+    },
+    {
+      "epoch": 1.0836054866100588,
+      "grad_norm": 0.2841418981552124,
+      "learning_rate": 4.03553178172417e-05,
+      "loss": 0.1151,
+      "step": 138
+    },
+    {
+      "epoch": 1.0914435009797518,
+      "grad_norm": 0.25916630029678345,
+      "learning_rate": 4.0173447515678916e-05,
+      "loss": 0.1178,
+      "step": 139
+    },
+    {
+      "epoch": 1.0992815153494448,
+      "grad_norm": 0.24618621170520782,
+      "learning_rate": 3.999029686399704e-05,
+      "loss": 0.1078,
+      "step": 140
+    },
+    {
+      "epoch": 1.1071195297191379,
+      "grad_norm": 0.3014475703239441,
+      "learning_rate": 3.9805881316624506e-05,
+      "loss": 0.1277,
+      "step": 141
+    },
+    {
+      "epoch": 1.1149575440888309,
+      "grad_norm": 0.2629724442958832,
+      "learning_rate": 3.962021643472284e-05,
+      "loss": 0.1168,
+      "step": 142
+    },
+    {
+      "epoch": 1.1227955584585239,
+      "grad_norm": 0.26206281781196594,
+      "learning_rate": 3.9433317884873664e-05,
+      "loss": 0.117,
+      "step": 143
+    },
+    {
+      "epoch": 1.1306335728282169,
+      "grad_norm": 0.2575000524520874,
+      "learning_rate": 3.9245201437756654e-05,
+      "loss": 0.1107,
+      "step": 144
+    },
+    {
+      "epoch": 1.1384715871979099,
+      "grad_norm": 0.2927393913269043,
+      "learning_rate": 3.905588296681886e-05,
+      "loss": 0.1216,
+      "step": 145
+    },
+    {
+      "epoch": 1.1463096015676029,
+      "grad_norm": 0.23872609436511993,
+      "learning_rate": 3.8865378446935216e-05,
+      "loss": 0.1067,
+      "step": 146
+    },
+    {
+      "epoch": 1.1541476159372959,
+      "grad_norm": 0.2771260738372803,
+      "learning_rate": 3.867370395306068e-05,
+      "loss": 0.1118,
+      "step": 147
+    },
+    {
+      "epoch": 1.1619856303069889,
+      "grad_norm": 0.2423640787601471,
+      "learning_rate": 3.848087565887365e-05,
+      "loss": 0.1179,
+      "step": 148
+    },
+    {
+      "epoch": 1.1698236446766819,
+      "grad_norm": 0.2449909895658493,
+      "learning_rate": 3.82869098354114e-05,
+      "loss": 0.1247,
+      "step": 149
+    },
+    {
+      "epoch": 1.1776616590463749,
+      "grad_norm": 0.2508186101913452,
+      "learning_rate": 3.8091822849696954e-05,
+      "loss": 0.1186,
+      "step": 150
+    },
+    {
+      "epoch": 1.1854996734160679,
+      "grad_norm": 0.22804874181747437,
+      "learning_rate": 3.7895631163358105e-05,
+      "loss": 0.1214,
+      "step": 151
+    },
+    {
+      "epoch": 1.1933376877857609,
+      "grad_norm": 0.23902134597301483,
+      "learning_rate": 3.769835133123836e-05,
+      "loss": 0.1118,
+      "step": 152
+    },
+    {
+      "epoch": 1.2011757021554539,
+      "grad_norm": 0.22123728692531586,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.1081,
+      "step": 153
+    },
+    {
+      "epoch": 1.2090137165251469,
+      "grad_norm": 0.22215275466442108,
+      "learning_rate": 3.7300593906719464e-05,
+      "loss": 0.1115,
+      "step": 154
+    },
+    {
+      "epoch": 1.21685173089484,
+      "grad_norm": 0.21011732518672943,
+      "learning_rate": 3.7100149877474974e-05,
+      "loss": 0.1105,
+      "step": 155
+    },
+    {
+      "epoch": 1.224689745264533,
+      "grad_norm": 0.25005996227264404,
+      "learning_rate": 3.689868482592684e-05,
+      "loss": 0.1143,
+      "step": 156
+    },
+    {
+      "epoch": 1.232527759634226,
+      "grad_norm": 0.22056043148040771,
+      "learning_rate": 3.66962157518902e-05,
+      "loss": 0.1113,
+      "step": 157
+    },
+    {
+      "epoch": 1.240365774003919,
+      "grad_norm": 0.226328045129776,
+      "learning_rate": 3.6492759739900564e-05,
+      "loss": 0.1153,
+      "step": 158
+    },
+    {
+      "epoch": 1.248203788373612,
+      "grad_norm": 0.21589796245098114,
+      "learning_rate": 3.628833395777224e-05,
+      "loss": 0.1034,
+      "step": 159
+    },
+    {
+      "epoch": 1.256041802743305,
+      "grad_norm": 0.2250795215368271,
+      "learning_rate": 3.608295565514965e-05,
+      "loss": 0.116,
+      "step": 160
+    },
+    {
+      "epoch": 1.263879817112998,
+      "grad_norm": 0.21993276476860046,
+      "learning_rate": 3.587664216205183e-05,
+      "loss": 0.1217,
+      "step": 161
+    },
+    {
+      "epoch": 1.2717178314826911,
+      "grad_norm": 0.21394269168376923,
+      "learning_rate": 3.5669410887410095e-05,
+      "loss": 0.1119,
+      "step": 162
+    },
+    {
+      "epoch": 1.2795558458523841,
+      "grad_norm": 0.20264342427253723,
+      "learning_rate": 3.546127931759903e-05,
+      "loss": 0.1076,
+      "step": 163
+    },
+    {
+      "epoch": 1.2873938602220771,
+      "grad_norm": 0.23118674755096436,
+      "learning_rate": 3.5252265014961006e-05,
+      "loss": 0.1189,
+      "step": 164
+    },
+    {
+      "epoch": 1.2952318745917701,
+      "grad_norm": 0.215042844414711,
+      "learning_rate": 3.504238561632424e-05,
+      "loss": 0.1096,
+      "step": 165
+    },
+    {
+      "epoch": 1.3030698889614631,
+      "grad_norm": 0.21117202937602997,
+      "learning_rate": 3.483165883151458e-05,
+      "loss": 0.1201,
+      "step": 166
+    },
+    {
+      "epoch": 1.3109079033311561,
+      "grad_norm": 0.21316753327846527,
+      "learning_rate": 3.4620102441861143e-05,
+      "loss": 0.1072,
+      "step": 167
+    },
+    {
+      "epoch": 1.3187459177008491,
+      "grad_norm": 0.21038144826889038,
+      "learning_rate": 3.4407734298695895e-05,
+      "loss": 0.1149,
+      "step": 168
+    },
+    {
+      "epoch": 1.3265839320705421,
+      "grad_norm": 0.2167028933763504,
+      "learning_rate": 3.4194572321847336e-05,
+      "loss": 0.1075,
+      "step": 169
+    },
+    {
+      "epoch": 1.3344219464402352,
+      "grad_norm": 0.21729543805122375,
+      "learning_rate": 3.398063449812844e-05,
+      "loss": 0.1056,
+      "step": 170
+    },
+    {
+      "epoch": 1.3422599608099282,
+      "grad_norm": 0.20175161957740784,
+      "learning_rate": 3.376593887981887e-05,
+      "loss": 0.1053,
+      "step": 171
+    },
+    {
+      "epoch": 1.3500979751796212,
+      "grad_norm": 0.21248085796833038,
+      "learning_rate": 3.355050358314172e-05,
+      "loss": 0.1059,
+      "step": 172
+    },
+    {
+      "epoch": 1.3579359895493142,
+      "grad_norm": 0.21043576300144196,
+      "learning_rate": 3.333434678673489e-05,
+      "loss": 0.1051,
+      "step": 173
+    },
+    {
+      "epoch": 1.3657740039190072,
+      "grad_norm": 0.21573218703269958,
+      "learning_rate": 3.311748673011709e-05,
+      "loss": 0.1168,
+      "step": 174
+    },
+    {
+      "epoch": 1.3736120182887002,
+      "grad_norm": 0.20607827603816986,
+      "learning_rate": 3.289994171214882e-05,
+      "loss": 0.1071,
+      "step": 175
+    },
+    {
+      "epoch": 1.3814500326583932,
+      "grad_norm": 0.20888157188892365,
+      "learning_rate": 3.268173008948826e-05,
+      "loss": 0.1066,
+      "step": 176
+    },
+    {
+      "epoch": 1.3892880470280862,
+      "grad_norm": 0.20718783140182495,
+      "learning_rate": 3.246287027504237e-05,
+      "loss": 0.1034,
+      "step": 177
+    },
+    {
+      "epoch": 1.3971260613977792,
+      "grad_norm": 0.21539223194122314,
+      "learning_rate": 3.224338073641312e-05,
+      "loss": 0.1284,
+      "step": 178
+    },
+    {
+      "epoch": 1.4049640757674722,
+      "grad_norm": 0.2096419632434845,
+      "learning_rate": 3.202327999433924e-05,
+      "loss": 0.108,
+      "step": 179
+    },
+    {
+      "epoch": 1.4128020901371652,
+      "grad_norm": 0.19958484172821045,
+      "learning_rate": 3.180258662113338e-05,
+      "loss": 0.1027,
+      "step": 180
+    },
+    {
+      "epoch": 1.4206401045068582,
+      "grad_norm": 0.20901361107826233,
+      "learning_rate": 3.158131923911498e-05,
+      "loss": 0.1025,
+      "step": 181
+    },
+    {
+      "epoch": 1.4284781188765514,
+      "grad_norm": 0.22095659375190735,
+      "learning_rate": 3.135949651903891e-05,
+      "loss": 0.1082,
+      "step": 182
+    },
+    {
+      "epoch": 1.4363161332462444,
+      "grad_norm": 0.20527766644954681,
+      "learning_rate": 3.1137137178519985e-05,
+      "loss": 0.1114,
+      "step": 183
+    },
+    {
+      "epoch": 1.4441541476159374,
+      "grad_norm": 0.21945808827877045,
+      "learning_rate": 3.091425998045356e-05,
+      "loss": 0.1178,
+      "step": 184
+    },
+    {
+      "epoch": 1.4519921619856304,
+      "grad_norm": 0.20016823709011078,
+      "learning_rate": 3.069088373143234e-05,
+      "loss": 0.1078,
+      "step": 185
+    },
+    {
+      "epoch": 1.4598301763553234,
+      "grad_norm": 0.20746341347694397,
+      "learning_rate": 3.04670272801594e-05,
+      "loss": 0.1134,
+      "step": 186
+    },
+    {
+      "epoch": 1.4676681907250164,
+      "grad_norm": 0.23339882493019104,
+      "learning_rate": 3.0242709515857758e-05,
+      "loss": 0.1176,
+      "step": 187
+    },
+    {
+      "epoch": 1.4755062050947094,
+      "grad_norm": 0.21998368203639984,
+      "learning_rate": 3.001794936667648e-05,
+      "loss": 0.118,
+      "step": 188
+    },
+    {
+      "epoch": 1.4833442194644024,
+      "grad_norm": 0.2124100923538208,
+      "learning_rate": 2.9792765798093465e-05,
+      "loss": 0.1028,
+      "step": 189
+    },
+    {
+      "epoch": 1.4911822338340954,
+      "grad_norm": 0.22203968465328217,
+      "learning_rate": 2.9567177811315178e-05,
+      "loss": 0.1127,
+      "step": 190
+    },
+    {
+      "epoch": 1.4990202482037884,
+      "grad_norm": 0.2147323042154312,
+      "learning_rate": 2.9341204441673266e-05,
+      "loss": 0.108,
+      "step": 191
+    },
+    {
+      "epoch": 1.5068582625734814,
+      "grad_norm": 0.22974567115306854,
+      "learning_rate": 2.9114864757018352e-05,
+      "loss": 0.1063,
+      "step": 192
+    },
+    {
+      "epoch": 1.5146962769431744,
+      "grad_norm": 0.21040025353431702,
+      "learning_rate": 2.8888177856111083e-05,
+      "loss": 0.1073,
+      "step": 193
+    },
+    {
+      "epoch": 1.5225342913128674,
+      "grad_norm": 0.20959196984767914,
+      "learning_rate": 2.8661162867010543e-05,
+      "loss": 0.0989,
+      "step": 194
+    },
+    {
+      "epoch": 1.5303723056825604,
+      "grad_norm": 0.2233189344406128,
+      "learning_rate": 2.8433838945460205e-05,
+      "loss": 0.1153,
+      "step": 195
+    },
+    {
+      "epoch": 1.5382103200522534,
+      "grad_norm": 0.21911266446113586,
+      "learning_rate": 2.820622527327158e-05,
+      "loss": 0.1052,
+      "step": 196
+    },
+    {
+      "epoch": 1.5460483344219464,
+      "grad_norm": 0.22261589765548706,
+      "learning_rate": 2.797834105670559e-05,
+      "loss": 0.1109,
+      "step": 197
+    },
+    {
+      "epoch": 1.5538863487916394,
+      "grad_norm": 0.19887444376945496,
+      "learning_rate": 2.7750205524851986e-05,
+      "loss": 0.1092,
+      "step": 198
+    },
+    {
+      "epoch": 1.5617243631613325,
+      "grad_norm": 0.20779380202293396,
+      "learning_rate": 2.752183792800671e-05,
+      "loss": 0.1076,
+      "step": 199
+    },
+    {
+      "epoch": 1.5695623775310255,
+      "grad_norm": 0.2083713710308075,
+      "learning_rate": 2.729325753604759e-05,
+      "loss": 0.1135,
+      "step": 200
+    },
+    {
+      "epoch": 1.5774003919007185,
+      "grad_norm": 0.2003626525402069,
+      "learning_rate": 2.7064483636808313e-05,
+      "loss": 0.0968,
+      "step": 201
+    },
+    {
+      "epoch": 1.5852384062704115,
+      "grad_norm": 0.20525719225406647,
+      "learning_rate": 2.6835535534450884e-05,
+      "loss": 0.1042,
+      "step": 202
+    },
+    {
+      "epoch": 1.5930764206401045,
+      "grad_norm": 0.21359968185424805,
+      "learning_rate": 2.6606432547836757e-05,
+      "loss": 0.1126,
+      "step": 203
+    },
+    {
+      "epoch": 1.6009144350097975,
+      "grad_norm": 0.2056104987859726,
+      "learning_rate": 2.6377194008896637e-05,
+      "loss": 0.1073,
+      "step": 204
+    },
+    {
+      "epoch": 1.6087524493794905,
+      "grad_norm": 0.19534572958946228,
+      "learning_rate": 2.6147839260999295e-05,
+      "loss": 0.113,
+      "step": 205
+    },
+    {
+      "epoch": 1.6165904637491835,
+      "grad_norm": 0.21906475722789764,
+      "learning_rate": 2.5918387657319316e-05,
+      "loss": 0.1098,
+      "step": 206
+    },
+    {
+      "epoch": 1.6244284781188765,
+      "grad_norm": 0.19928069412708282,
+      "learning_rate": 2.5688858559204053e-05,
+      "loss": 0.1059,
+      "step": 207
+    },
+    {
+      "epoch": 1.6322664924885695,
+      "grad_norm": 0.19612881541252136,
+      "learning_rate": 2.5459271334539934e-05,
+      "loss": 0.0951,
+      "step": 208
+    },
+    {
+      "epoch": 1.6401045068582625,
+      "grad_norm": 0.2001960426568985,
+      "learning_rate": 2.5229645356118163e-05,
+      "loss": 0.1039,
+      "step": 209
+    },
+    {
+      "epoch": 1.6479425212279555,
+      "grad_norm": 0.20145660638809204,
+      "learning_rate": 2.5e-05,
+      "loss": 0.1049,
+      "step": 210
+    },
+    {
+      "epoch": 1.6557805355976485,
+      "grad_norm": 0.18295256793498993,
+      "learning_rate": 2.4770354643881843e-05,
+      "loss": 0.0985,
+      "step": 211
+    },
+    {
+      "epoch": 1.6636185499673415,
+      "grad_norm": 0.2013918161392212,
+      "learning_rate": 2.4540728665460065e-05,
+      "loss": 0.1065,
+      "step": 212
+    },
+    {
+      "epoch": 1.6714565643370345,
+      "grad_norm": 0.20308274030685425,
+      "learning_rate": 2.4311141440795953e-05,
+      "loss": 0.1205,
+      "step": 213
+    },
+    {
+      "epoch": 1.6792945787067275,
+      "grad_norm": 0.19719456136226654,
+      "learning_rate": 2.4081612342680694e-05,
+      "loss": 0.1027,
+      "step": 214
+    },
+    {
+      "epoch": 1.6871325930764205,
+      "grad_norm": 0.19175922870635986,
+      "learning_rate": 2.3852160739000707e-05,
+      "loss": 0.1,
+      "step": 215
+    },
+    {
+      "epoch": 1.6949706074461135,
+      "grad_norm": 0.20407816767692566,
+      "learning_rate": 2.3622805991103362e-05,
+      "loss": 0.102,
+      "step": 216
+    },
+    {
+      "epoch": 1.7028086218158065,
+      "grad_norm": 0.20707584917545319,
+      "learning_rate": 2.3393567452163252e-05,
+      "loss": 0.1139,
+      "step": 217
+    },
+    {
+      "epoch": 1.7106466361854997,
+      "grad_norm": 0.18942616879940033,
+      "learning_rate": 2.3164464465549118e-05,
+      "loss": 0.091,
+      "step": 218
+    },
+    {
+      "epoch": 1.7184846505551927,
+      "grad_norm": 0.2033768594264984,
+      "learning_rate": 2.2935516363191693e-05,
+      "loss": 0.1093,
+      "step": 219
+    },
+    {
+      "epoch": 1.7263226649248857,
+      "grad_norm": 0.2145160436630249,
+      "learning_rate": 2.270674246395241e-05,
+      "loss": 0.1083,
+      "step": 220
+    },
+    {
+      "epoch": 1.7341606792945787,
+      "grad_norm": 0.19836893677711487,
+      "learning_rate": 2.2478162071993298e-05,
+      "loss": 0.1011,
+      "step": 221
+    },
+    {
+      "epoch": 1.7419986936642717,
+      "grad_norm": 0.20721815526485443,
+      "learning_rate": 2.224979447514802e-05,
+      "loss": 0.0975,
+      "step": 222
+    },
+    {
+      "epoch": 1.7498367080339647,
+      "grad_norm": 0.20113082230091095,
+      "learning_rate": 2.202165894329441e-05,
+      "loss": 0.1086,
+      "step": 223
+    },
+    {
+      "epoch": 1.7576747224036577,
+      "grad_norm": 0.19542041420936584,
+      "learning_rate": 2.179377472672842e-05,
+      "loss": 0.1029,
+      "step": 224
+    },
+    {
+      "epoch": 1.7655127367733507,
+      "grad_norm": 0.19994117319583893,
+      "learning_rate": 2.1566161054539798e-05,
+      "loss": 0.1071,
+      "step": 225
+    },
+    {
+      "epoch": 1.7733507511430437,
+      "grad_norm": 0.20020049810409546,
+      "learning_rate": 2.1338837132989466e-05,
+      "loss": 0.0995,
+      "step": 226
+    },
+    {
+      "epoch": 1.7811887655127367,
+      "grad_norm": 0.2069658637046814,
+      "learning_rate": 2.111182214388893e-05,
+      "loss": 0.1077,
+      "step": 227
+    },
+    {
+      "epoch": 1.7890267798824298,
+      "grad_norm": 0.21685247123241425,
+      "learning_rate": 2.088513524298165e-05,
+      "loss": 0.1084,
+      "step": 228
+    },
+    {
+      "epoch": 1.7968647942521228,
+      "grad_norm": 0.19922901690006256,
+      "learning_rate": 2.0658795558326743e-05,
+      "loss": 0.0999,
+      "step": 229
+    },
+    {
+      "epoch": 1.8047028086218158,
+      "grad_norm": 0.20437341928482056,
+      "learning_rate": 2.043282218868483e-05,
+      "loss": 0.1076,
+      "step": 230
+    },
+    {
+      "epoch": 1.812540822991509,
+      "grad_norm": 0.19708654284477234,
+      "learning_rate": 2.0207234201906547e-05,
+      "loss": 0.0971,
+      "step": 231
+    },
+    {
+      "epoch": 1.820378837361202,
+      "grad_norm": 0.21140620112419128,
+      "learning_rate": 1.9982050633323522e-05,
+      "loss": 0.1046,
+      "step": 232
+    },
+    {
+      "epoch": 1.828216851730895,
+      "grad_norm": 0.19234000146389008,
+      "learning_rate": 1.9757290484142244e-05,
+      "loss": 0.0957,
+      "step": 233
+    },
+    {
+      "epoch": 1.836054866100588,
+      "grad_norm": 0.2157324254512787,
+      "learning_rate": 1.9532972719840607e-05,
+      "loss": 0.1082,
+      "step": 234
+    },
+    {
+      "epoch": 1.843892880470281,
+      "grad_norm": 0.19539855420589447,
+      "learning_rate": 1.9309116268567674e-05,
+      "loss": 0.1014,
+      "step": 235
+    },
+    {
+      "epoch": 1.851730894839974,
+      "grad_norm": 0.1897420883178711,
+      "learning_rate": 1.908574001954644e-05,
+      "loss": 0.094,
+      "step": 236
+    },
+    {
+      "epoch": 1.859568909209667,
+      "grad_norm": 0.20616742968559265,
+      "learning_rate": 1.8862862821480025e-05,
+      "loss": 0.1118,
+      "step": 237
+    },
+    {
+      "epoch": 1.86740692357936,
+      "grad_norm": 0.1957971155643463,
+      "learning_rate": 1.864050348096109e-05,
+      "loss": 0.0948,
+      "step": 238
+    },
+    {
+      "epoch": 1.875244937949053,
+      "grad_norm": 0.18928657472133636,
+      "learning_rate": 1.8418680760885027e-05,
+      "loss": 0.0963,
+      "step": 239
+    },
+    {
+      "epoch": 1.883082952318746,
+      "grad_norm": 0.20447275042533875,
+      "learning_rate": 1.819741337886662e-05,
+      "loss": 0.096,
+      "step": 240
+    },
+    {
+      "epoch": 1.890920966688439,
+      "grad_norm": 0.19518423080444336,
+      "learning_rate": 1.797672000566077e-05,
+      "loss": 0.1042,
+      "step": 241
+    },
+    {
+      "epoch": 1.898758981058132,
+      "grad_norm": 0.18361036479473114,
+      "learning_rate": 1.775661926358689e-05,
+      "loss": 0.0975,
+      "step": 242
+    },
+    {
+      "epoch": 1.906596995427825,
+      "grad_norm": 0.19658760726451874,
+      "learning_rate": 1.7537129724957642e-05,
+      "loss": 0.1063,
+      "step": 243
+    },
+    {
+      "epoch": 1.914435009797518,
+      "grad_norm": 0.194893017411232,
+      "learning_rate": 1.7318269910511736e-05,
+      "loss": 0.1007,
+      "step": 244
+    },
+    {
+      "epoch": 1.922273024167211,
+      "grad_norm": 0.19551898539066315,
+      "learning_rate": 1.710005828785119e-05,
+      "loss": 0.0989,
+      "step": 245
+    },
+    {
+      "epoch": 1.930111038536904,
+      "grad_norm": 0.19419115781784058,
+      "learning_rate": 1.6882513269882917e-05,
+      "loss": 0.0894,
+      "step": 246
+    },
+    {
+      "epoch": 1.937949052906597,
+      "grad_norm": 0.19363057613372803,
+      "learning_rate": 1.666565321326512e-05,
+      "loss": 0.0954,
+      "step": 247
+    },
+    {
+      "epoch": 1.94578706727629,
+      "grad_norm": 0.18952466547489166,
+      "learning_rate": 1.6449496416858284e-05,
+      "loss": 0.0906,
+      "step": 248
+    },
+    {
+      "epoch": 1.953625081645983,
+      "grad_norm": 0.19003801047801971,
+      "learning_rate": 1.6234061120181142e-05,
+      "loss": 0.0982,
+      "step": 249
+    },
+    {
+      "epoch": 1.961463096015676,
+      "grad_norm": 0.18380698561668396,
+      "learning_rate": 1.601936550187157e-05,
+      "loss": 0.099,
+      "step": 250
+    },
+    {
+      "epoch": 1.969301110385369,
+      "grad_norm": 0.1997416615486145,
+      "learning_rate": 1.5805427678152677e-05,
+      "loss": 0.1027,
+      "step": 251
+    },
+    {
+      "epoch": 1.977139124755062,
+      "grad_norm": 0.18836817145347595,
+      "learning_rate": 1.5592265701304114e-05,
+      "loss": 0.0983,
+      "step": 252
+    },
+    {
+      "epoch": 1.984977139124755,
+      "grad_norm": 0.2013559639453888,
+      "learning_rate": 1.5379897558138862e-05,
+      "loss": 0.102,
+      "step": 253
+    },
+    {
+      "epoch": 1.992815153494448,
+      "grad_norm": 0.19445432722568512,
+      "learning_rate": 1.5168341168485423e-05,
+      "loss": 0.0953,
+      "step": 254
+    },
+    {
+      "epoch": 2.002612671456564,
+      "grad_norm": 0.3252091705799103,
+      "learning_rate": 1.495761438367577e-05,
+      "loss": 0.1459,
+      "step": 255
+    },
+    {
+      "epoch": 2.010450685826257,
+      "grad_norm": 0.20789092779159546,
+      "learning_rate": 1.4747734985039e-05,
+      "loss": 0.0344,
+      "step": 256
+    },
+    {
+      "epoch": 2.01828870019595,
+      "grad_norm": 0.17120763659477234,
+      "learning_rate": 1.4538720682400969e-05,
+      "loss": 0.0347,
+      "step": 257
+    },
+    {
+      "epoch": 2.026126714565643,
+      "grad_norm": 0.14296689629554749,
+      "learning_rate": 1.433058911258991e-05,
+      "loss": 0.0296,
+      "step": 258
+    },
+    {
+      "epoch": 2.033964728935336,
+      "grad_norm": 0.18135866522789001,
+      "learning_rate": 1.4123357837948175e-05,
+      "loss": 0.0332,
+      "step": 259
+    },
+    {
+      "epoch": 2.041802743305029,
+      "grad_norm": 0.20270270109176636,
+      "learning_rate": 1.3917044344850355e-05,
+      "loss": 0.0322,
+      "step": 260
+    },
+    {
+      "epoch": 2.049640757674722,
+      "grad_norm": 0.1781613826751709,
+      "learning_rate": 1.3711666042227772e-05,
+      "loss": 0.0309,
+      "step": 261
+    },
+    {
+      "epoch": 2.057478772044415,
+      "grad_norm": 0.17559246718883514,
+      "learning_rate": 1.350724026009944e-05,
+      "loss": 0.0333,
+      "step": 262
+    },
+    {
+      "epoch": 2.065316786414108,
+      "grad_norm": 0.14222192764282227,
+      "learning_rate": 1.330378424810981e-05,
+      "loss": 0.0309,
+      "step": 263
+    },
+    {
+      "epoch": 2.0731548007838017,
+      "grad_norm": 0.15568028390407562,
+      "learning_rate": 1.3101315174073162e-05,
+      "loss": 0.0324,
+      "step": 264
+    },
+    {
+      "epoch": 2.0809928151534947,
+      "grad_norm": 0.15495392680168152,
+      "learning_rate": 1.2899850122525037e-05,
+      "loss": 0.0329,
+      "step": 265
+    },
+    {
+      "epoch": 2.0888308295231877,
+      "grad_norm": 0.16096574068069458,
+      "learning_rate": 1.2699406093280547e-05,
+      "loss": 0.0317,
+      "step": 266
+    },
+    {
+      "epoch": 2.0966688438928807,
+      "grad_norm": 0.13694337010383606,
+      "learning_rate": 1.2500000000000006e-05,
+      "loss": 0.0282,
+      "step": 267
+    },
+    {
+      "epoch": 2.1045068582625737,
+      "grad_norm": 0.13896653056144714,
+      "learning_rate": 1.2301648668761647e-05,
+      "loss": 0.0289,
+      "step": 268
+    },
+    {
+      "epoch": 2.1123448726322667,
+      "grad_norm": 0.15309906005859375,
+      "learning_rate": 1.2104368836641908e-05,
+      "loss": 0.0321,
+      "step": 269
+    },
+    {
+      "epoch": 2.1201828870019597,
+      "grad_norm": 0.16871266067028046,
+      "learning_rate": 1.1908177150303055e-05,
+      "loss": 0.0335,
+      "step": 270
+    },
+    {
+      "epoch": 2.1280209013716527,
+      "grad_norm": 0.15097489953041077,
+      "learning_rate": 1.1713090164588607e-05,
+      "loss": 0.0302,
+      "step": 271
+    },
+    {
+      "epoch": 2.1358589157413457,
+      "grad_norm": 0.1583205759525299,
+      "learning_rate": 1.1519124341126346e-05,
+      "loss": 0.0337,
+      "step": 272
+    },
+    {
+      "epoch": 2.1436969301110387,
+      "grad_norm": 0.1436772346496582,
+      "learning_rate": 1.1326296046939333e-05,
+      "loss": 0.0296,
+      "step": 273
+    },
+    {
+      "epoch": 2.1515349444807317,
+      "grad_norm": 0.1444602757692337,
+      "learning_rate": 1.113462155306478e-05,
+      "loss": 0.0296,
+      "step": 274
+    },
+    {
+      "epoch": 2.1593729588504247,
+      "grad_norm": 0.13352134823799133,
+      "learning_rate": 1.0944117033181151e-05,
+      "loss": 0.0287,
+      "step": 275
+    },
+    {
+      "epoch": 2.1672109732201177,
+      "grad_norm": 0.1418055146932602,
+      "learning_rate": 1.0754798562243345e-05,
+      "loss": 0.0294,
+      "step": 276
+    },
+    {
+      "epoch": 2.1750489875898107,
+      "grad_norm": 0.12928378582000732,
+      "learning_rate": 1.0566682115126344e-05,
+      "loss": 0.0274,
+      "step": 277
+    },
+    {
+      "epoch": 2.1828870019595037,
+      "grad_norm": 0.13986484706401825,
+      "learning_rate": 1.037978356527716e-05,
+      "loss": 0.0328,
+      "step": 278
+    },
+    {
+      "epoch": 2.1907250163291967,
+      "grad_norm": 0.13403278589248657,
+      "learning_rate": 1.0194118683375503e-05,
+      "loss": 0.0296,
+      "step": 279
+    },
+    {
+      "epoch": 2.1985630306988897,
+      "grad_norm": 0.14188171923160553,
+      "learning_rate": 1.0009703136002957e-05,
+      "loss": 0.0299,
+      "step": 280
+    },
+    {
+      "epoch": 2.2064010450685827,
+      "grad_norm": 0.13912995159626007,
+      "learning_rate": 9.826552484321087e-06,
+      "loss": 0.03,
+      "step": 281
+    },
+    {
+      "epoch": 2.2142390594382757,
+      "grad_norm": 0.14268967509269714,
+      "learning_rate": 9.644682182758306e-06,
+      "loss": 0.0293,
+      "step": 282
+    },
+    {
+      "epoch": 2.2220770738079687,
+      "grad_norm": 0.14812886714935303,
+      "learning_rate": 9.464107577705886e-06,
+      "loss": 0.0269,
+      "step": 283
+    },
+    {
+      "epoch": 2.2299150881776617,
+      "grad_norm": 0.1425481140613556,
+      "learning_rate": 9.284843906222948e-06,
+      "loss": 0.0274,
+      "step": 284
+    },
+    {
+      "epoch": 2.2377531025473547,
+      "grad_norm": 0.13893434405326843,
+      "learning_rate": 9.106906294750805e-06,
+      "loss": 0.0292,
+      "step": 285
+    },
+    {
+      "epoch": 2.2455911169170477,
+      "grad_norm": 0.13851140439510345,
+      "learning_rate": 8.930309757836517e-06,
+      "loss": 0.0321,
+      "step": 286
+    },
+    {
+      "epoch": 2.2534291312867407,
+      "grad_norm": 0.13696281611919403,
+      "learning_rate": 8.755069196866014e-06,
+      "loss": 0.0286,
+      "step": 287
+    },
+    {
+      "epoch": 2.2612671456564337,
+      "grad_norm": 0.13719762861728668,
+      "learning_rate": 8.581199398806641e-06,
+      "loss": 0.0314,
+      "step": 288
+    },
+    {
+      "epoch": 2.2691051600261267,
+      "grad_norm": 0.1380627155303955,
+      "learning_rate": 8.40871503495947e-06,
+      "loss": 0.0309,
+      "step": 289
+    },
+    {
+      "epoch": 2.2769431743958197,
+      "grad_norm": 0.13105949759483337,
+      "learning_rate": 8.237630659721277e-06,
+      "loss": 0.0259,
+      "step": 290
+    },
+    {
+      "epoch": 2.2847811887655127,
+      "grad_norm": 0.1306370496749878,
+      "learning_rate": 8.067960709356478e-06,
+      "loss": 0.0264,
+      "step": 291
+    },
+    {
+      "epoch": 2.2926192031352057,
+      "grad_norm": 0.13892598450183868,
+      "learning_rate": 7.899719500778923e-06,
+      "loss": 0.0308,
+      "step": 292
+    },
+    {
+      "epoch": 2.3004572175048987,
+      "grad_norm": 0.13053514063358307,
+      "learning_rate": 7.732921230343892e-06,
+      "loss": 0.0276,
+      "step": 293
+    },
+    {
+      "epoch": 2.3082952318745917,
+      "grad_norm": 0.1372438371181488,
+      "learning_rate": 7.5675799726501155e-06,
+      "loss": 0.0271,
+      "step": 294
+    },
+    {
+      "epoch": 2.3161332462442847,
+      "grad_norm": 0.1358174979686737,
+      "learning_rate": 7.403709679352217e-06,
+      "loss": 0.027,
+      "step": 295
+    },
+    {
+      "epoch": 2.3239712606139777,
+      "grad_norm": 0.13224904239177704,
+      "learning_rate": 7.2413241779834e-06,
+      "loss": 0.027,
+      "step": 296
+    },
+    {
+      "epoch": 2.3318092749836707,
+      "grad_norm": 0.13056500256061554,
+      "learning_rate": 7.080437170788723e-06,
+      "loss": 0.0282,
+      "step": 297
+    },
+    {
+      "epoch": 2.3396472893533637,
+      "grad_norm": 0.13113614916801453,
+      "learning_rate": 6.921062233568831e-06,
+      "loss": 0.028,
+      "step": 298
+    },
+    {
+      "epoch": 2.3474853037230567,
+      "grad_norm": 0.13239531219005585,
+      "learning_rate": 6.763212814534484e-06,
+      "loss": 0.0272,
+      "step": 299
+    },
+    {
+      "epoch": 2.3553233180927498,
+      "grad_norm": 0.12943539023399353,
+      "learning_rate": 6.606902233171711e-06,
+      "loss": 0.0247,
+      "step": 300
+    },
+    {
+      "epoch": 2.3631613324624428,
+      "grad_norm": 0.12735775113105774,
+      "learning_rate": 6.452143679117964e-06,
+      "loss": 0.0263,
+      "step": 301
+    },
+    {
+      "epoch": 2.3709993468321358,
+      "grad_norm": 0.13229192793369293,
+      "learning_rate": 6.298950211049115e-06,
+      "loss": 0.0283,
+      "step": 302
+    },
+    {
+      "epoch": 2.3788373612018288,
+      "grad_norm": 0.12901271879673004,
+      "learning_rate": 6.147334755577596e-06,
+      "loss": 0.0258,
+      "step": 303
+    },
+    {
+      "epoch": 2.3866753755715218,
+      "grad_norm": 0.13257119059562683,
+      "learning_rate": 5.997310106161589e-06,
+      "loss": 0.0279,
+      "step": 304
+    },
+    {
+      "epoch": 2.3945133899412148,
+      "grad_norm": 0.12502893805503845,
+      "learning_rate": 5.848888922025553e-06,
+      "loss": 0.0259,
+      "step": 305
+    },
+    {
+      "epoch": 2.4023514043109078,
+      "grad_norm": 0.13395585119724274,
+      "learning_rate": 5.702083727091978e-06,
+      "loss": 0.0286,
+      "step": 306
+    },
+    {
+      "epoch": 2.4101894186806008,
+      "grad_norm": 0.12500248849391937,
+      "learning_rate": 5.556906908924655e-06,
+      "loss": 0.0266,
+      "step": 307
+    },
+    {
+      "epoch": 2.4180274330502938,
+      "grad_norm": 0.13586992025375366,
+      "learning_rate": 5.413370717683347e-06,
+      "loss": 0.0302,
+      "step": 308
+    },
+    {
+      "epoch": 2.4258654474199868,
+      "grad_norm": 0.13480009138584137,
+      "learning_rate": 5.271487265090163e-06,
+      "loss": 0.0281,
+      "step": 309
+    },
+    {
+      "epoch": 2.43370346178968,
+      "grad_norm": 0.14151740074157715,
+      "learning_rate": 5.131268523407509e-06,
+      "loss": 0.0291,
+      "step": 310
+    },
+    {
+      "epoch": 2.441541476159373,
+      "grad_norm": 0.13139696419239044,
+      "learning_rate": 4.992726324427901e-06,
+      "loss": 0.0283,
+      "step": 311
+    },
+    {
+      "epoch": 2.449379490529066,
+      "grad_norm": 0.12287548929452896,
+      "learning_rate": 4.855872358475547e-06,
+      "loss": 0.0243,
+      "step": 312
+    },
+    {
+      "epoch": 2.4572175048987592,
+      "grad_norm": 0.12606504559516907,
+      "learning_rate": 4.720718173419947e-06,
+      "loss": 0.0271,
+      "step": 313
+    },
+    {
+      "epoch": 2.465055519268452,
+      "grad_norm": 0.1324881911277771,
+      "learning_rate": 4.587275173701428e-06,
+      "loss": 0.0281,
+      "step": 314
+    },
+    {
+      "epoch": 2.4728935336381452,
+      "grad_norm": 0.13066086173057556,
+      "learning_rate": 4.4555546193688735e-06,
+      "loss": 0.029,
+      "step": 315
+    },
+    {
+      "epoch": 2.480731548007838,
+      "grad_norm": 0.12259192019701004,
+      "learning_rate": 4.3255676251295456e-06,
+      "loss": 0.0248,
+      "step": 316
+    },
+    {
+      "epoch": 2.4885695623775312,
+      "grad_norm": 0.13243158161640167,
+      "learning_rate": 4.19732515941125e-06,
+      "loss": 0.0282,
+      "step": 317
+    },
+    {
+      "epoch": 2.496407576747224,
+      "grad_norm": 0.13366110622882843,
+      "learning_rate": 4.070838043436786e-06,
+      "loss": 0.0295,
+      "step": 318
+    },
+    {
+      "epoch": 2.5042455911169172,
+      "grad_norm": 0.1202232763171196,
+      "learning_rate": 3.94611695031086e-06,
+      "loss": 0.0241,
+      "step": 319
+    },
+    {
+      "epoch": 2.51208360548661,
+      "grad_norm": 0.13794013857841492,
+      "learning_rate": 3.8231724041194515e-06,
+      "loss": 0.0284,
+      "step": 320
+    },
+    {
+      "epoch": 2.5199216198563033,
+      "grad_norm": 0.13122816383838654,
+      "learning_rate": 3.7020147790418263e-06,
+      "loss": 0.0261,
+      "step": 321
+    },
+    {
+      "epoch": 2.527759634225996,
+      "grad_norm": 0.12636157870292664,
+      "learning_rate": 3.582654298475091e-06,
+      "loss": 0.024,
+      "step": 322
+    },
+    {
+      "epoch": 2.5355976485956893,
+      "grad_norm": 0.12592478096485138,
+      "learning_rate": 3.4651010341716028e-06,
+      "loss": 0.0261,
+      "step": 323
+    },
+    {
+      "epoch": 2.5434356629653823,
+      "grad_norm": 0.1321692168712616,
+      "learning_rate": 3.3493649053890326e-06,
+      "loss": 0.0255,
+      "step": 324
+    },
+    {
+      "epoch": 2.5512736773350753,
+      "grad_norm": 0.12444434314966202,
+      "learning_rate": 3.2354556780534424e-06,
+      "loss": 0.0243,
+      "step": 325
+    },
+    {
+      "epoch": 2.5591116917047683,
+      "grad_norm": 0.1310444325208664,
+      "learning_rate": 3.1233829639351563e-06,
+      "loss": 0.0264,
+      "step": 326
+    },
+    {
+      "epoch": 2.5669497060744613,
+      "grad_norm": 0.13162964582443237,
+      "learning_rate": 3.013156219837776e-06,
+      "loss": 0.0266,
+      "step": 327
+    },
+    {
+      "epoch": 2.5747877204441543,
+      "grad_norm": 0.12017495930194855,
+      "learning_rate": 2.9047847468001404e-06,
+      "loss": 0.0248,
+      "step": 328
+    },
+    {
+      "epoch": 2.5826257348138473,
+      "grad_norm": 0.13093486428260803,
+      "learning_rate": 2.7982776893115627e-06,
+      "loss": 0.0279,
+      "step": 329
+    },
+    {
+      "epoch": 2.5904637491835403,
+      "grad_norm": 0.12869121134281158,
+      "learning_rate": 2.6936440345401493e-06,
+      "loss": 0.0252,
+      "step": 330
+    },
+    {
+      "epoch": 2.5983017635532333,
+      "grad_norm": 0.1348036676645279,
+      "learning_rate": 2.5908926115744997e-06,
+      "loss": 0.0281,
+      "step": 331
+    },
+    {
+      "epoch": 2.6061397779229263,
+      "grad_norm": 0.12896090745925903,
+      "learning_rate": 2.4900320906786593e-06,
+      "loss": 0.0267,
+      "step": 332
+    },
+    {
+      "epoch": 2.6139777922926193,
+      "grad_norm": 0.1246756836771965,
+      "learning_rate": 2.391070982560564e-06,
+      "loss": 0.0256,
+      "step": 333
+    },
+    {
+      "epoch": 2.6218158066623123,
+      "grad_norm": 0.12264394015073776,
+      "learning_rate": 2.2940176376538445e-06,
+      "loss": 0.0235,
+      "step": 334
+    },
+    {
+      "epoch": 2.6296538210320053,
+      "grad_norm": 0.12561777234077454,
+      "learning_rate": 2.1988802454132403e-06,
+      "loss": 0.025,
+      "step": 335
+    },
+    {
+      "epoch": 2.6374918354016983,
+      "grad_norm": 0.11878997832536697,
+      "learning_rate": 2.1056668336235622e-06,
+      "loss": 0.0247,
+      "step": 336
+    },
+    {
+      "epoch": 2.6453298497713913,
+      "grad_norm": 0.12836302816867828,
+      "learning_rate": 2.0143852677223075e-06,
+      "loss": 0.0285,
+      "step": 337
+    },
+    {
+      "epoch": 2.6531678641410843,
+      "grad_norm": 0.12017780542373657,
+      "learning_rate": 1.9250432501359354e-06,
+      "loss": 0.025,
+      "step": 338
+    },
+    {
+      "epoch": 2.6610058785107773,
+      "grad_norm": 0.11886154115200043,
+      "learning_rate": 1.837648319629956e-06,
+      "loss": 0.0244,
+      "step": 339
+    },
+    {
+      "epoch": 2.6688438928804703,
+      "grad_norm": 0.13172321021556854,
+      "learning_rate": 1.7522078506728074e-06,
+      "loss": 0.0269,
+      "step": 340
+    },
+    {
+      "epoch": 2.6766819072501633,
+      "grad_norm": 0.12933161854743958,
+      "learning_rate": 1.6687290528135723e-06,
+      "loss": 0.027,
+      "step": 341
+    },
+    {
+      "epoch": 2.6845199216198563,
+      "grad_norm": 0.13184285163879395,
+      "learning_rate": 1.5872189700736339e-06,
+      "loss": 0.0266,
+      "step": 342
+    },
+    {
+      "epoch": 2.6923579359895493,
+      "grad_norm": 0.12135521322488785,
+      "learning_rate": 1.5076844803522922e-06,
+      "loss": 0.0236,
+      "step": 343
+    },
+    {
+      "epoch": 2.7001959503592423,
+      "grad_norm": 0.12548145651817322,
+      "learning_rate": 1.4301322948464147e-06,
+      "loss": 0.0259,
+      "step": 344
+    },
+    {
+      "epoch": 2.7080339647289353,
+      "grad_norm": 0.12587113678455353,
+      "learning_rate": 1.3545689574841342e-06,
+      "loss": 0.0282,
+      "step": 345
+    },
+    {
+      "epoch": 2.7158719790986283,
+      "grad_norm": 0.12320797145366669,
+      "learning_rate": 1.2810008443726456e-06,
+      "loss": 0.0251,
+      "step": 346
+    },
+    {
+      "epoch": 2.7237099934683213,
+      "grad_norm": 0.12896938621997833,
+      "learning_rate": 1.2094341632602064e-06,
+      "loss": 0.0258,
+      "step": 347
+    },
+    {
+      "epoch": 2.7315480078380143,
+      "grad_norm": 0.12183412909507751,
+      "learning_rate": 1.1398749530123127e-06,
+      "loss": 0.0243,
+      "step": 348
+    },
+    {
+      "epoch": 2.7393860222077073,
+      "grad_norm": 0.12513582408428192,
+      "learning_rate": 1.0723290831021471e-06,
+      "loss": 0.0254,
+      "step": 349
+    },
+    {
+      "epoch": 2.7472240365774003,
+      "grad_norm": 0.13293574750423431,
+      "learning_rate": 1.006802253115277e-06,
+      "loss": 0.0287,
+      "step": 350
+    },
+    {
+      "epoch": 2.7550620509470933,
+      "grad_norm": 0.11619393527507782,
+      "learning_rate": 9.432999922687396e-07,
+      "loss": 0.0229,
+      "step": 351
+    },
+    {
+      "epoch": 2.7629000653167863,
+      "grad_norm": 0.12378672510385513,
+      "learning_rate": 8.818276589444896e-07,
+      "loss": 0.0247,
+      "step": 352
+    },
+    {
+      "epoch": 2.7707380796864793,
+      "grad_norm": 0.1200721487402916,
+      "learning_rate": 8.223904402372334e-07,
+      "loss": 0.0244,
+      "step": 353
+    },
+    {
+      "epoch": 2.7785760940561723,
+      "grad_norm": 0.11517145484685898,
+      "learning_rate": 7.649933515167407e-07,
+      "loss": 0.0228,
+      "step": 354
+    },
+    {
+      "epoch": 2.7864141084258653,
+      "grad_norm": 0.12590059638023376,
+      "learning_rate": 7.096412360046545e-07,
+      "loss": 0.0257,
+      "step": 355
+    },
+    {
+      "epoch": 2.7942521227955583,
+      "grad_norm": 0.12404926866292953,
+      "learning_rate": 6.563387643658076e-07,
+      "loss": 0.0252,
+      "step": 356
+    },
+    {
+      "epoch": 2.8020901371652513,
+      "grad_norm": 0.12552441656589508,
+      "learning_rate": 6.050904343141095e-07,
+      "loss": 0.026,
+      "step": 357
+    },
+    {
+      "epoch": 2.8099281515349444,
+      "grad_norm": 0.12259657680988312,
+      "learning_rate": 5.5590057023302e-07,
+      "loss": 0.0254,
+      "step": 358
+    },
+    {
+      "epoch": 2.8177661659046374,
+      "grad_norm": 0.11798401176929474,
+      "learning_rate": 5.087733228106517e-07,
+      "loss": 0.024,
+      "step": 359
+    },
+    {
+      "epoch": 2.8256041802743304,
+      "grad_norm": 0.12366917729377747,
+      "learning_rate": 4.637126686895532e-07,
+      "loss": 0.0257,
+      "step": 360
+    },
+    {
+      "epoch": 2.8334421946440234,
+      "grad_norm": 0.1320623904466629,
+      "learning_rate": 4.207224101311247e-07,
+      "loss": 0.0277,
+      "step": 361
+    },
+    {
+      "epoch": 2.8412802090137164,
+      "grad_norm": 0.11438736319541931,
+      "learning_rate": 3.7980617469479953e-07,
+      "loss": 0.0214,
+      "step": 362
+    },
+    {
+      "epoch": 2.8491182233834094,
+      "grad_norm": 0.11790072917938232,
+      "learning_rate": 3.4096741493194197e-07,
+      "loss": 0.0238,
+      "step": 363
+    },
+    {
+      "epoch": 2.856956237753103,
+      "grad_norm": 0.12108506262302399,
+      "learning_rate": 3.0420940809451624e-07,
+      "loss": 0.0241,
+      "step": 364
+    },
+    {
+      "epoch": 2.8647942521227954,
+      "grad_norm": 0.12375527620315552,
+      "learning_rate": 2.6953525585855234e-07,
+      "loss": 0.0251,
+      "step": 365
+    },
+    {
+      "epoch": 2.872632266492489,
+      "grad_norm": 0.12097339332103729,
+      "learning_rate": 2.3694788406241896e-07,
+      "loss": 0.025,
+      "step": 366
+    },
+    {
+      "epoch": 2.8804702808621814,
+      "grad_norm": 0.12990431487560272,
+      "learning_rate": 2.064500424599436e-07,
+      "loss": 0.0257,
+      "step": 367
+    },
+    {
+      "epoch": 2.888308295231875,
+      "grad_norm": 0.11965355277061462,
+      "learning_rate": 1.7804430448837839e-07,
+      "loss": 0.0267,
+      "step": 368
+    },
+    {
+      "epoch": 2.8961463096015674,
+      "grad_norm": 0.12657780945301056,
+      "learning_rate": 1.517330670512629e-07,
+      "loss": 0.0257,
+      "step": 369
+    },
+    {
+      "epoch": 2.903984323971261,
+      "grad_norm": 0.12143310904502869,
+      "learning_rate": 1.27518550316158e-07,
+      "loss": 0.0242,
+      "step": 370
+    },
+    {
+      "epoch": 2.9118223383409534,
+      "grad_norm": 0.16820454597473145,
+      "learning_rate": 1.0540279752731253e-07,
+      "loss": 0.0241,
+      "step": 371
+    },
+    {
+      "epoch": 2.919660352710647,
+      "grad_norm": 0.12241283059120178,
+      "learning_rate": 8.538767483325383e-08,
+      "loss": 0.0235,
+      "step": 372
+    },
+    {
+      "epoch": 2.9274983670803394,
+      "grad_norm": 0.13118112087249756,
+      "learning_rate": 6.747487112931661e-08,
+      "loss": 0.0281,
+      "step": 373
+    },
+    {
+      "epoch": 2.935336381450033,
+      "grad_norm": 0.11441458761692047,
+      "learning_rate": 5.166589791513465e-08,
+      "loss": 0.023,
+      "step": 374
+    },
+    {
+      "epoch": 2.9431743958197254,
+      "grad_norm": 0.12388182431459427,
+      "learning_rate": 3.796208916709565e-08,
+      "loss": 0.0266,
+      "step": 375
+    },
+    {
+      "epoch": 2.951012410189419,
+      "grad_norm": 0.12555594742298126,
+      "learning_rate": 2.636460122578399e-08,
+      "loss": 0.0236,
+      "step": 376
+    },
+    {
+      "epoch": 2.958850424559112,
+      "grad_norm": 0.1175137609243393,
+      "learning_rate": 1.6874412698408836e-08,
+      "loss": 0.0221,
+      "step": 377
+    },
+    {
+      "epoch": 2.966688438928805,
+      "grad_norm": 0.12019700556993484,
+      "learning_rate": 9.492324376214612e-09,
+      "loss": 0.0237,
+      "step": 378
+    },
+    {
+      "epoch": 2.974526453298498,
+      "grad_norm": 0.12358499318361282,
+      "learning_rate": 4.218959166932268e-09,
+      "loss": 0.0248,
+      "step": 379
+    },
+    {
+      "epoch": 2.982364467668191,
+      "grad_norm": 0.13017946481704712,
+      "learning_rate": 1.0547620421907934e-09,
+      "loss": 0.0284,
+      "step": 380
+    },
+    {
+      "epoch": 2.990202482037884,
+      "grad_norm": 0.11907056719064713,
+      "learning_rate": 0.0,
+      "loss": 0.0251,
+      "step": 381
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 381,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1405397234876416.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:056928e64622f52e940b8e226aff799b3dc0463ba24813d484ba8fb3352778fc
+size 7608

zero_to_fp32.py

@@ -0,0 +1,674 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example:
+#   python zero_to_fp32.py . output_dir/
+#   or
+#   python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import json
+from tqdm import tqdm
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+    total_files = len(files)
+    state_dicts = []
+    for f in files:
+        state_dict = torch.load(f, map_location=device)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    if zero_stage <= 2:
+        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    elif zero_stage == 3:
+        # if there is more than one param group, there will be multiple flattened tensors - one
+        # flattened tensor per group - for simplicity merge them into a single tensor
+        #
+        # XXX: could make the script more memory efficient for when there are multiple groups - it
+        # will require matching the sub-lists of param_shapes for each param group flattened tensor
+
+        fp32_flat_groups = [
+            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
+        ]
+
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = fp32_flat_groups[0].numel() * world_size
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    for name, shape in tqdm(param_shapes.items(), desc='Gathering Sharded Weights'):
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # XXX: memory usage doubles here
+        state_dict[name] = torch.cat(
+            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
+            0).narrow(0, 0, unpartitioned_numel).view(shape)
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+                                               output_dir,
+                                               max_shard_size="5GB",
+                                               safe_serialization=False,
+                                               tag=None,
+                                               exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_dir``: directory to the pytorch fp32 state_dict output files
+        - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+        - ``safe_serialization``:  whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+    # Dependency pre-check
+    if safe_serialization:
+        try:
+            from safetensors.torch import save_file
+        except ImportError:
+            print('If you want to use `safe_serialization`, please `pip install safetensors`')
+            raise
+    if max_shard_size is not None:
+        try:
+            from huggingface_hub import split_torch_state_dict_into_shards
+        except ImportError:
+            print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+            raise
+
+    # Convert zero checkpoint to state_dict
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters)
+
+    # Shard the model if it is too big.
+    weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+    if max_shard_size is not None:
+        filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+        state_dict_split = split_torch_state_dict_into_shards(state_dict,
+                                                              filename_pattern=filename_pattern,
+                                                              max_shard_size=max_shard_size)
+    else:
+        from collections import namedtuple
+        StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+        state_dict_split = StateDictSplit(is_sharded=False,
+                                          filename_to_tensors={weights_name: list(state_dict.keys())})
+
+    # Save the model
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard = {tensor: state_dict[tensor].contiguous() for tensor in tensors}
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard, output_path)
+
+    # Save index if sharded
+    if state_dict_split.is_sharded:
+        index = {
+            "metadata": state_dict_split.metadata,
+            "weight_map": state_dict_split.tensor_to_filename,
+        }
+        save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+        save_index_file = os.path.join(output_dir, save_index_file)
+        with open(save_index_file, "w", encoding="utf-8") as f:
+            content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+            f.write(content)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument("output_dir",
+                        type=str,
+                        help="directory to the pytorch fp32 state_dict output files"
+                        "(e.g. path/checkpoint-12-output/)")
+    parser.add_argument(
+        "--max_shard_size",
+        type=str,
+        default="5GB",
+        help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+        "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+        "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+        "without CPU OOM issues.")
+    parser.add_argument(
+        "--safe_serialization",
+        default=False,
+        action='store_true',
+        help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_dir,
+                                               max_shard_size=args.max_shard_size,
+                                               safe_serialization=args.safe_serialization,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)