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 +1,28 @@
+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

chat_template.jinja

@@ -0,0 +1,89 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# 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>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\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" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if message.content is string %}
+        {%- set content = message.content %}
+    {%- else %}
+        {%- set content = '' %}
+    {%- endif %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

config.json

@@ -0,0 +1,68 @@
+{
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2560,
+  "initializer_range": 0.02,
+  "intermediate_size": 9728,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 40960,
+  "max_window_layers": 36,
+  "model_type": "qwen3",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 8,
+  "pad_token_id": 151643,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "4.57.1",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

generation_config.json

@@ -0,0 +1,12 @@
+{
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.6,
+  "top_k": 20,
+  "top_p": 0.95,
+  "transformers_version": "4.57.1"
+}

global_step5000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4e534c2e89f1a0d5bcee3455252c9480cdf83d22a23ff2a6151f534eac1367b
+size 6033707356

global_step5000/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ebba6f4cd68ae94b09047bdfd63fbe6f34c346de0f6e37c3f36608cbfd2fb35
+size 6033707356

global_step5000/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6097768c0a5310ca5c4caa0cb90fcae05e82d170bee57f36147761d6c0ffe2ac
+size 6033707356

global_step5000/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07ea2ba60ca9da048b110209f3ca6951f981c8196cd8351121b5cca2b49b5af8
+size 6033707356

global_step5000/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d495398b9e91f7444bbe4921472ba7a9950ddb9b000187a7f45365928319589
+size 6033707356

global_step5000/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0ed5ffd7b25273d3afe0178018700f8ae8836f58a7dea68d609bc2a459b939d
+size 6033707356

global_step5000/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:925954336d40a8b778c98751d44cdc4a67c2f7e90c4db66e11372f2f523fe6c5
+size 6033707356

global_step5000/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8894942795c7516144e92fb5fc0e666c61af867ccc523b88209ce0cf2e9dfb36
+size 6033707356

global_step5000/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3935c45c06d9cdbeb07ff736eabdb713f015eb0317ca3724c0b1b97ae9ff697
+size 202052

global_step5000/zero_pp_rank_1_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b83e9031c1377b1d6c8c5d2383b14fb29485462d097cdffb2adcdc42b0174460
+size 201988

global_step5000/zero_pp_rank_2_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b74678b1756dd099b81b8e359b31009621253119885ebda8616057cf7f298c9
+size 201988

global_step5000/zero_pp_rank_3_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c011c9e2ce6a66da942715a7fa5bd6042bf4578bcd139ebf50bd01b4784719b7
+size 201988

global_step5000/zero_pp_rank_4_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7eaa1c831075bebc0171bde8a5c8a6f6f1734ff9f27457f4f698b13047fbebb3
+size 201988

global_step5000/zero_pp_rank_5_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c97760695e9f032c9bb832df1fe979fdad66d8110debcc43e78a7349f47b871
+size 201988

global_step5000/zero_pp_rank_6_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdd3a79edc514e5fdfb7907b4d8b7b2bea49a5efb38cb2b58b17256f14dbcd24
+size 201988

global_step5000/zero_pp_rank_7_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26690f4b16f00a8e761e2044333d1f448624e7e87d8e7f63c037fbab3dd5cf3b
+size 201988

latest

@@ -0,0 +1 @@
+global_step5000

model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:afe210f7d0eb174827f27db8c55c07f39110d3e52c4087af5eb5355d0d8e5042
+size 4967215360

model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e14b95c2f8e40743f379af0303188c5169570b176786672ae36260d2a2dc3148
+size 3077766632

model.safetensors.index.json

@@ -0,0 +1,406 @@
+{
+  "metadata": {
+    "total_parameters": 196096,
+    "total_size": 8044936192
+  },
+  "weight_map": {
+    "model.embed_tokens.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.20.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.norm.weight": "model-00002-of-00002.safetensors"
+  }
+}

rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:955b895101e13c81ba510512313a06782795770a0bf998c90f718166d25f1664
+size 15984

rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:096e351ace65ff91008be171a45173ed6369cc639fce73a288f671041e24b0ec
+size 15984

rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f21c61b1a7e793bbdec183de3b52da90042305234bc7e5887986655cd3fc2192
+size 15984

rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:348742498d81780506d9760c655a7a7555185b5fbd70a7ae296d88fd9aeecd84
+size 15984

rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:399c4700dab9ae7b754110ce307fb7e26e22cb49b5bb233c435f0f12b77c202f
+size 15984

rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:864ea2379cc907eb4189c52706cb978150d9c26e18abf74679590729a8f0c8e8
+size 15984

rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25755ba07299ee6ff45936dd04df329596319c9f8095af71e6f3a219e7543e26
+size 15984

rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:908f018cd701ed629c41299726da4a25f202f20a1d4bc2075a2266ed4013db3a
+size 15984

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56ef787b02e3b40a13b599800ad87036df33f559cd1634b3c4473f4a4dc2298e
+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": "<|im_end|>",
+    "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:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
+size 11422654

tokenizer_config.json

@@ -0,0 +1,240 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151665": {
+      "content": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

trainer_state.json

@@ -0,0 +1,3534 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.03999968000255998,
+  "eval_steps": 500,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 7.999936000511996e-05,
+      "grad_norm": 4.414881453815827,
+      "learning_rate": 7.199424046076314e-09,
+      "loss": 0.5441,
+      "step": 10
+    },
+    {
+      "epoch": 0.00015999872001023991,
+      "grad_norm": 4.004957247533114,
+      "learning_rate": 1.519878409727222e-08,
+      "loss": 0.5353,
+      "step": 20
+    },
+    {
+      "epoch": 0.00023999808001535987,
+      "grad_norm": 4.889790729834478,
+      "learning_rate": 2.3198144148468124e-08,
+      "loss": 0.537,
+      "step": 30
+    },
+    {
+      "epoch": 0.00031999744002047983,
+      "grad_norm": 3.7138301639437508,
+      "learning_rate": 3.119750419966403e-08,
+      "loss": 0.5488,
+      "step": 40
+    },
+    {
+      "epoch": 0.0003999968000255998,
+      "grad_norm": 4.208028943214464,
+      "learning_rate": 3.919686425085993e-08,
+      "loss": 0.5557,
+      "step": 50
+    },
+    {
+      "epoch": 0.00047999616003071974,
+      "grad_norm": 4.326896533998126,
+      "learning_rate": 4.719622430205584e-08,
+      "loss": 0.5061,
+      "step": 60
+    },
+    {
+      "epoch": 0.0005599955200358397,
+      "grad_norm": 7.1723877700496415,
+      "learning_rate": 5.519558435325175e-08,
+      "loss": 0.5193,
+      "step": 70
+    },
+    {
+      "epoch": 0.0006399948800409597,
+      "grad_norm": 3.892238785520058,
+      "learning_rate": 6.319494440444764e-08,
+      "loss": 0.5311,
+      "step": 80
+    },
+    {
+      "epoch": 0.0007199942400460796,
+      "grad_norm": 3.4517472540795513,
+      "learning_rate": 7.119430445564356e-08,
+      "loss": 0.5232,
+      "step": 90
+    },
+    {
+      "epoch": 0.0007999936000511996,
+      "grad_norm": 4.094057558667446,
+      "learning_rate": 7.919366450683946e-08,
+      "loss": 0.5225,
+      "step": 100
+    },
+    {
+      "epoch": 0.0008799929600563195,
+      "grad_norm": 3.4395682865399144,
+      "learning_rate": 8.719302455803536e-08,
+      "loss": 0.5371,
+      "step": 110
+    },
+    {
+      "epoch": 0.0009599923200614395,
+      "grad_norm": 3.7029712275218682,
+      "learning_rate": 9.519238460923127e-08,
+      "loss": 0.5305,
+      "step": 120
+    },
+    {
+      "epoch": 0.0010399916800665594,
+      "grad_norm": 3.493027269180126,
+      "learning_rate": 1.0319174466042718e-07,
+      "loss": 0.5167,
+      "step": 130
+    },
+    {
+      "epoch": 0.0011199910400716794,
+      "grad_norm": 3.480790725115363,
+      "learning_rate": 1.1119110471162308e-07,
+      "loss": 0.5543,
+      "step": 140
+    },
+    {
+      "epoch": 0.0011999904000767993,
+      "grad_norm": 3.0350778792458564,
+      "learning_rate": 1.1919046476281897e-07,
+      "loss": 0.5188,
+      "step": 150
+    },
+    {
+      "epoch": 0.0012799897600819193,
+      "grad_norm": 2.7703962024693216,
+      "learning_rate": 1.271898248140149e-07,
+      "loss": 0.5176,
+      "step": 160
+    },
+    {
+      "epoch": 0.0013599891200870393,
+      "grad_norm": 2.7312812150288277,
+      "learning_rate": 1.351891848652108e-07,
+      "loss": 0.5206,
+      "step": 170
+    },
+    {
+      "epoch": 0.0014399884800921593,
+      "grad_norm": 2.6772442085163837,
+      "learning_rate": 1.431885449164067e-07,
+      "loss": 0.5187,
+      "step": 180
+    },
+    {
+      "epoch": 0.0015199878400972793,
+      "grad_norm": 2.6017371604025503,
+      "learning_rate": 1.5118790496760262e-07,
+      "loss": 0.4858,
+      "step": 190
+    },
+    {
+      "epoch": 0.0015999872001023993,
+      "grad_norm": 2.361921882047865,
+      "learning_rate": 1.5918726501879854e-07,
+      "loss": 0.4965,
+      "step": 200
+    },
+    {
+      "epoch": 0.0016799865601075192,
+      "grad_norm": 1.9949603945902301,
+      "learning_rate": 1.6718662506999443e-07,
+      "loss": 0.49,
+      "step": 210
+    },
+    {
+      "epoch": 0.001759985920112639,
+      "grad_norm": 1.689017850332906,
+      "learning_rate": 1.7518598512119031e-07,
+      "loss": 0.4474,
+      "step": 220
+    },
+    {
+      "epoch": 0.001839985280117759,
+      "grad_norm": 1.7526108484730667,
+      "learning_rate": 1.8318534517238623e-07,
+      "loss": 0.4758,
+      "step": 230
+    },
+    {
+      "epoch": 0.001919984640122879,
+      "grad_norm": 1.6619250196852287,
+      "learning_rate": 1.9118470522358212e-07,
+      "loss": 0.4551,
+      "step": 240
+    },
+    {
+      "epoch": 0.001999984000127999,
+      "grad_norm": 1.4244411983222107,
+      "learning_rate": 1.9918406527477803e-07,
+      "loss": 0.4666,
+      "step": 250
+    },
+    {
+      "epoch": 0.0020799833601331187,
+      "grad_norm": 1.4091371713945773,
+      "learning_rate": 2.0718342532597392e-07,
+      "loss": 0.4744,
+      "step": 260
+    },
+    {
+      "epoch": 0.002159982720138239,
+      "grad_norm": 1.4037698231377602,
+      "learning_rate": 2.1518278537716986e-07,
+      "loss": 0.4577,
+      "step": 270
+    },
+    {
+      "epoch": 0.0022399820801433587,
+      "grad_norm": 1.1233973407058262,
+      "learning_rate": 2.2318214542836575e-07,
+      "loss": 0.4442,
+      "step": 280
+    },
+    {
+      "epoch": 0.002319981440148479,
+      "grad_norm": 1.0934034827825632,
+      "learning_rate": 2.3118150547956164e-07,
+      "loss": 0.4746,
+      "step": 290
+    },
+    {
+      "epoch": 0.0023999808001535987,
+      "grad_norm": 0.9547032856231036,
+      "learning_rate": 2.3918086553075753e-07,
+      "loss": 0.4631,
+      "step": 300
+    },
+    {
+      "epoch": 0.002479980160158719,
+      "grad_norm": 1.011447249716743,
+      "learning_rate": 2.4718022558195345e-07,
+      "loss": 0.4359,
+      "step": 310
+    },
+    {
+      "epoch": 0.0025599795201638386,
+      "grad_norm": 0.7506687829382205,
+      "learning_rate": 2.5517958563314936e-07,
+      "loss": 0.4554,
+      "step": 320
+    },
+    {
+      "epoch": 0.002639978880168959,
+      "grad_norm": 0.8268141903301145,
+      "learning_rate": 2.631789456843453e-07,
+      "loss": 0.4444,
+      "step": 330
+    },
+    {
+      "epoch": 0.0027199782401740786,
+      "grad_norm": 0.9120933866995852,
+      "learning_rate": 2.711783057355412e-07,
+      "loss": 0.4662,
+      "step": 340
+    },
+    {
+      "epoch": 0.0027999776001791984,
+      "grad_norm": 0.9481570868182263,
+      "learning_rate": 2.791776657867371e-07,
+      "loss": 0.4677,
+      "step": 350
+    },
+    {
+      "epoch": 0.0028799769601843186,
+      "grad_norm": 0.8010445040150771,
+      "learning_rate": 2.8717702583793297e-07,
+      "loss": 0.4196,
+      "step": 360
+    },
+    {
+      "epoch": 0.0029599763201894383,
+      "grad_norm": 0.8670839473253033,
+      "learning_rate": 2.951763858891289e-07,
+      "loss": 0.4492,
+      "step": 370
+    },
+    {
+      "epoch": 0.0030399756801945585,
+      "grad_norm": 0.8536232676044045,
+      "learning_rate": 3.031757459403248e-07,
+      "loss": 0.4375,
+      "step": 380
+    },
+    {
+      "epoch": 0.0031199750401996783,
+      "grad_norm": 0.8135677817452803,
+      "learning_rate": 3.111751059915207e-07,
+      "loss": 0.438,
+      "step": 390
+    },
+    {
+      "epoch": 0.0031999744002047985,
+      "grad_norm": 0.9667962048008838,
+      "learning_rate": 3.191744660427166e-07,
+      "loss": 0.4348,
+      "step": 400
+    },
+    {
+      "epoch": 0.0032799737602099183,
+      "grad_norm": 0.885059264680228,
+      "learning_rate": 3.271738260939125e-07,
+      "loss": 0.4295,
+      "step": 410
+    },
+    {
+      "epoch": 0.0033599731202150385,
+      "grad_norm": 0.7112007856245484,
+      "learning_rate": 3.3517318614510846e-07,
+      "loss": 0.4412,
+      "step": 420
+    },
+    {
+      "epoch": 0.0034399724802201583,
+      "grad_norm": 0.9486242885147141,
+      "learning_rate": 3.431725461963043e-07,
+      "loss": 0.4357,
+      "step": 430
+    },
+    {
+      "epoch": 0.003519971840225278,
+      "grad_norm": 0.7792187180449996,
+      "learning_rate": 3.5117190624750024e-07,
+      "loss": 0.4053,
+      "step": 440
+    },
+    {
+      "epoch": 0.0035999712002303982,
+      "grad_norm": 0.8013118911013118,
+      "learning_rate": 3.591712662986961e-07,
+      "loss": 0.4526,
+      "step": 450
+    },
+    {
+      "epoch": 0.003679970560235518,
+      "grad_norm": 0.9290353407648725,
+      "learning_rate": 3.6717062634989207e-07,
+      "loss": 0.4394,
+      "step": 460
+    },
+    {
+      "epoch": 0.003759969920240638,
+      "grad_norm": 0.7440275824040646,
+      "learning_rate": 3.7516998640108793e-07,
+      "loss": 0.4354,
+      "step": 470
+    },
+    {
+      "epoch": 0.003839969280245758,
+      "grad_norm": 0.6661735557442742,
+      "learning_rate": 3.8316934645228385e-07,
+      "loss": 0.4132,
+      "step": 480
+    },
+    {
+      "epoch": 0.003919968640250878,
+      "grad_norm": 0.80714967397643,
+      "learning_rate": 3.9116870650347976e-07,
+      "loss": 0.4247,
+      "step": 490
+    },
+    {
+      "epoch": 0.003999968000255998,
+      "grad_norm": 0.7269691108846945,
+      "learning_rate": 3.9916806655467563e-07,
+      "loss": 0.4308,
+      "step": 500
+    },
+    {
+      "epoch": 0.004079967360261118,
+      "grad_norm": 0.8112092339496537,
+      "learning_rate": 4.071674266058716e-07,
+      "loss": 0.4215,
+      "step": 510
+    },
+    {
+      "epoch": 0.0041599667202662375,
+      "grad_norm": 0.8465738727092779,
+      "learning_rate": 4.1516678665706746e-07,
+      "loss": 0.4438,
+      "step": 520
+    },
+    {
+      "epoch": 0.004239966080271358,
+      "grad_norm": 0.7947114246478345,
+      "learning_rate": 4.2316614670826337e-07,
+      "loss": 0.402,
+      "step": 530
+    },
+    {
+      "epoch": 0.004319965440276478,
+      "grad_norm": 0.771445835897062,
+      "learning_rate": 4.311655067594593e-07,
+      "loss": 0.4434,
+      "step": 540
+    },
+    {
+      "epoch": 0.004399964800281598,
+      "grad_norm": 0.7143580000434352,
+      "learning_rate": 4.3916486681065515e-07,
+      "loss": 0.4227,
+      "step": 550
+    },
+    {
+      "epoch": 0.004479964160286717,
+      "grad_norm": 0.7487773075552359,
+      "learning_rate": 4.471642268618511e-07,
+      "loss": 0.438,
+      "step": 560
+    },
+    {
+      "epoch": 0.004559963520291838,
+      "grad_norm": 0.7373810653849879,
+      "learning_rate": 4.55163586913047e-07,
+      "loss": 0.4288,
+      "step": 570
+    },
+    {
+      "epoch": 0.004639962880296958,
+      "grad_norm": 0.7104505393064982,
+      "learning_rate": 4.631629469642429e-07,
+      "loss": 0.4462,
+      "step": 580
+    },
+    {
+      "epoch": 0.004719962240302078,
+      "grad_norm": 0.6996674322090365,
+      "learning_rate": 4.7116230701543876e-07,
+      "loss": 0.4263,
+      "step": 590
+    },
+    {
+      "epoch": 0.004799961600307197,
+      "grad_norm": 0.779468888606517,
+      "learning_rate": 4.791616670666347e-07,
+      "loss": 0.4378,
+      "step": 600
+    },
+    {
+      "epoch": 0.0048799609603123175,
+      "grad_norm": 0.6348456379271163,
+      "learning_rate": 4.871610271178306e-07,
+      "loss": 0.4098,
+      "step": 610
+    },
+    {
+      "epoch": 0.004959960320317438,
+      "grad_norm": 0.6384341588537313,
+      "learning_rate": 4.951603871690265e-07,
+      "loss": 0.4489,
+      "step": 620
+    },
+    {
+      "epoch": 0.005039959680322557,
+      "grad_norm": 0.6779051066808497,
+      "learning_rate": 5.031597472202224e-07,
+      "loss": 0.4522,
+      "step": 630
+    },
+    {
+      "epoch": 0.005119959040327677,
+      "grad_norm": 0.7405337420145796,
+      "learning_rate": 5.111591072714183e-07,
+      "loss": 0.4248,
+      "step": 640
+    },
+    {
+      "epoch": 0.0051999584003327975,
+      "grad_norm": 0.6915688533513078,
+      "learning_rate": 5.191584673226143e-07,
+      "loss": 0.4129,
+      "step": 650
+    },
+    {
+      "epoch": 0.005279957760337918,
+      "grad_norm": 0.6044444399034763,
+      "learning_rate": 5.271578273738101e-07,
+      "loss": 0.4237,
+      "step": 660
+    },
+    {
+      "epoch": 0.005359957120343037,
+      "grad_norm": 0.7401624173362278,
+      "learning_rate": 5.351571874250061e-07,
+      "loss": 0.4216,
+      "step": 670
+    },
+    {
+      "epoch": 0.005439956480348157,
+      "grad_norm": 0.7496141168534413,
+      "learning_rate": 5.43156547476202e-07,
+      "loss": 0.4268,
+      "step": 680
+    },
+    {
+      "epoch": 0.005519955840353277,
+      "grad_norm": 0.730614366570662,
+      "learning_rate": 5.511559075273978e-07,
+      "loss": 0.4589,
+      "step": 690
+    },
+    {
+      "epoch": 0.005599955200358397,
+      "grad_norm": 0.7836211351094068,
+      "learning_rate": 5.591552675785937e-07,
+      "loss": 0.4176,
+      "step": 700
+    },
+    {
+      "epoch": 0.005679954560363517,
+      "grad_norm": 0.6648189022606215,
+      "learning_rate": 5.671546276297896e-07,
+      "loss": 0.4274,
+      "step": 710
+    },
+    {
+      "epoch": 0.005759953920368637,
+      "grad_norm": 0.6607901995290275,
+      "learning_rate": 5.751539876809856e-07,
+      "loss": 0.428,
+      "step": 720
+    },
+    {
+      "epoch": 0.005839953280373757,
+      "grad_norm": 0.8422142881250906,
+      "learning_rate": 5.831533477321815e-07,
+      "loss": 0.4206,
+      "step": 730
+    },
+    {
+      "epoch": 0.005919952640378877,
+      "grad_norm": 0.8636103630308042,
+      "learning_rate": 5.911527077833774e-07,
+      "loss": 0.4384,
+      "step": 740
+    },
+    {
+      "epoch": 0.005999952000383997,
+      "grad_norm": 0.7020283376881055,
+      "learning_rate": 5.991520678345733e-07,
+      "loss": 0.4336,
+      "step": 750
+    },
+    {
+      "epoch": 0.006079951360389117,
+      "grad_norm": 0.6637768636661737,
+      "learning_rate": 6.071514278857692e-07,
+      "loss": 0.421,
+      "step": 760
+    },
+    {
+      "epoch": 0.006159950720394236,
+      "grad_norm": 0.6720228370970702,
+      "learning_rate": 6.151507879369651e-07,
+      "loss": 0.4425,
+      "step": 770
+    },
+    {
+      "epoch": 0.006239950080399357,
+      "grad_norm": 0.7915308024009126,
+      "learning_rate": 6.231501479881609e-07,
+      "loss": 0.4665,
+      "step": 780
+    },
+    {
+      "epoch": 0.006319949440404477,
+      "grad_norm": 0.7005483141808596,
+      "learning_rate": 6.311495080393569e-07,
+      "loss": 0.4371,
+      "step": 790
+    },
+    {
+      "epoch": 0.006399948800409597,
+      "grad_norm": 0.7085842808188052,
+      "learning_rate": 6.391488680905528e-07,
+      "loss": 0.4481,
+      "step": 800
+    },
+    {
+      "epoch": 0.006479948160414716,
+      "grad_norm": 0.797364113205394,
+      "learning_rate": 6.471482281417488e-07,
+      "loss": 0.43,
+      "step": 810
+    },
+    {
+      "epoch": 0.0065599475204198366,
+      "grad_norm": 0.7769920944189029,
+      "learning_rate": 6.551475881929446e-07,
+      "loss": 0.4421,
+      "step": 820
+    },
+    {
+      "epoch": 0.006639946880424957,
+      "grad_norm": 0.7190098449382926,
+      "learning_rate": 6.631469482441405e-07,
+      "loss": 0.4524,
+      "step": 830
+    },
+    {
+      "epoch": 0.006719946240430077,
+      "grad_norm": 0.8999089940382301,
+      "learning_rate": 6.711463082953363e-07,
+      "loss": 0.4327,
+      "step": 840
+    },
+    {
+      "epoch": 0.006799945600435196,
+      "grad_norm": 0.8071258672281636,
+      "learning_rate": 6.791456683465323e-07,
+      "loss": 0.3999,
+      "step": 850
+    },
+    {
+      "epoch": 0.0068799449604403165,
+      "grad_norm": 0.9718244104757909,
+      "learning_rate": 6.871450283977283e-07,
+      "loss": 0.4109,
+      "step": 860
+    },
+    {
+      "epoch": 0.006959944320445437,
+      "grad_norm": 0.7260805899057289,
+      "learning_rate": 6.951443884489241e-07,
+      "loss": 0.4254,
+      "step": 870
+    },
+    {
+      "epoch": 0.007039943680450556,
+      "grad_norm": 0.6793848246836304,
+      "learning_rate": 7.031437485001201e-07,
+      "loss": 0.4168,
+      "step": 880
+    },
+    {
+      "epoch": 0.007119943040455676,
+      "grad_norm": 0.8220405330990818,
+      "learning_rate": 7.11143108551316e-07,
+      "loss": 0.4368,
+      "step": 890
+    },
+    {
+      "epoch": 0.0071999424004607964,
+      "grad_norm": 0.8650021092848164,
+      "learning_rate": 7.191424686025118e-07,
+      "loss": 0.4336,
+      "step": 900
+    },
+    {
+      "epoch": 0.007279941760465917,
+      "grad_norm": 0.6667689577948444,
+      "learning_rate": 7.271418286537078e-07,
+      "loss": 0.4325,
+      "step": 910
+    },
+    {
+      "epoch": 0.007359941120471036,
+      "grad_norm": 0.6443211014509481,
+      "learning_rate": 7.351411887049036e-07,
+      "loss": 0.443,
+      "step": 920
+    },
+    {
+      "epoch": 0.007439940480476156,
+      "grad_norm": 0.6067913066251965,
+      "learning_rate": 7.431405487560996e-07,
+      "loss": 0.4095,
+      "step": 930
+    },
+    {
+      "epoch": 0.007519939840481276,
+      "grad_norm": 0.6935512372810877,
+      "learning_rate": 7.511399088072954e-07,
+      "loss": 0.4272,
+      "step": 940
+    },
+    {
+      "epoch": 0.007599939200486396,
+      "grad_norm": 0.6599113657719949,
+      "learning_rate": 7.591392688584914e-07,
+      "loss": 0.3942,
+      "step": 950
+    },
+    {
+      "epoch": 0.007679938560491516,
+      "grad_norm": 0.611225655896161,
+      "learning_rate": 7.671386289096873e-07,
+      "loss": 0.4451,
+      "step": 960
+    },
+    {
+      "epoch": 0.007759937920496636,
+      "grad_norm": 0.757015652343119,
+      "learning_rate": 7.751379889608831e-07,
+      "loss": 0.4059,
+      "step": 970
+    },
+    {
+      "epoch": 0.007839937280501755,
+      "grad_norm": 0.8456384517410107,
+      "learning_rate": 7.831373490120791e-07,
+      "loss": 0.4278,
+      "step": 980
+    },
+    {
+      "epoch": 0.007919936640506876,
+      "grad_norm": 0.7996016208371923,
+      "learning_rate": 7.911367090632751e-07,
+      "loss": 0.4536,
+      "step": 990
+    },
+    {
+      "epoch": 0.007999936000511996,
+      "grad_norm": 0.6228736248515305,
+      "learning_rate": 7.991360691144709e-07,
+      "loss": 0.3969,
+      "step": 1000
+    },
+    {
+      "epoch": 0.008079935360517116,
+      "grad_norm": 0.6440105424946402,
+      "learning_rate": 8.071354291656668e-07,
+      "loss": 0.4307,
+      "step": 1010
+    },
+    {
+      "epoch": 0.008159934720522236,
+      "grad_norm": 0.6636370756757088,
+      "learning_rate": 8.151347892168628e-07,
+      "loss": 0.4019,
+      "step": 1020
+    },
+    {
+      "epoch": 0.008239934080527356,
+      "grad_norm": 0.6534728964626215,
+      "learning_rate": 8.231341492680586e-07,
+      "loss": 0.4357,
+      "step": 1030
+    },
+    {
+      "epoch": 0.008319933440532475,
+      "grad_norm": 0.6854238349847609,
+      "learning_rate": 8.311335093192545e-07,
+      "loss": 0.4156,
+      "step": 1040
+    },
+    {
+      "epoch": 0.008399932800537595,
+      "grad_norm": 0.7692797686595098,
+      "learning_rate": 8.391328693704504e-07,
+      "loss": 0.4364,
+      "step": 1050
+    },
+    {
+      "epoch": 0.008479932160542715,
+      "grad_norm": 0.6892139807862941,
+      "learning_rate": 8.471322294216464e-07,
+      "loss": 0.4595,
+      "step": 1060
+    },
+    {
+      "epoch": 0.008559931520547836,
+      "grad_norm": 0.6786670330601587,
+      "learning_rate": 8.551315894728423e-07,
+      "loss": 0.4207,
+      "step": 1070
+    },
+    {
+      "epoch": 0.008639930880552956,
+      "grad_norm": 0.6118366771253826,
+      "learning_rate": 8.631309495240381e-07,
+      "loss": 0.4186,
+      "step": 1080
+    },
+    {
+      "epoch": 0.008719930240558076,
+      "grad_norm": 0.7157718863255984,
+      "learning_rate": 8.711303095752341e-07,
+      "loss": 0.4412,
+      "step": 1090
+    },
+    {
+      "epoch": 0.008799929600563196,
+      "grad_norm": 0.7102096103999658,
+      "learning_rate": 8.791296696264299e-07,
+      "loss": 0.4205,
+      "step": 1100
+    },
+    {
+      "epoch": 0.008879928960568315,
+      "grad_norm": 0.6326552697725677,
+      "learning_rate": 8.871290296776258e-07,
+      "loss": 0.4161,
+      "step": 1110
+    },
+    {
+      "epoch": 0.008959928320573435,
+      "grad_norm": 0.7455469536196851,
+      "learning_rate": 8.951283897288219e-07,
+      "loss": 0.4259,
+      "step": 1120
+    },
+    {
+      "epoch": 0.009039927680578555,
+      "grad_norm": 0.8219540514047248,
+      "learning_rate": 9.031277497800177e-07,
+      "loss": 0.4423,
+      "step": 1130
+    },
+    {
+      "epoch": 0.009119927040583675,
+      "grad_norm": 0.6428846698552027,
+      "learning_rate": 9.111271098312136e-07,
+      "loss": 0.432,
+      "step": 1140
+    },
+    {
+      "epoch": 0.009199926400588795,
+      "grad_norm": 0.7959015617833209,
+      "learning_rate": 9.191264698824094e-07,
+      "loss": 0.4374,
+      "step": 1150
+    },
+    {
+      "epoch": 0.009279925760593916,
+      "grad_norm": 0.722690696817927,
+      "learning_rate": 9.271258299336054e-07,
+      "loss": 0.4177,
+      "step": 1160
+    },
+    {
+      "epoch": 0.009359925120599036,
+      "grad_norm": 0.7583146312532495,
+      "learning_rate": 9.351251899848013e-07,
+      "loss": 0.4349,
+      "step": 1170
+    },
+    {
+      "epoch": 0.009439924480604156,
+      "grad_norm": 0.8841341873240007,
+      "learning_rate": 9.431245500359971e-07,
+      "loss": 0.4323,
+      "step": 1180
+    },
+    {
+      "epoch": 0.009519923840609274,
+      "grad_norm": 0.6844990736142492,
+      "learning_rate": 9.511239100871932e-07,
+      "loss": 0.4195,
+      "step": 1190
+    },
+    {
+      "epoch": 0.009599923200614395,
+      "grad_norm": 0.7129122863481299,
+      "learning_rate": 9.59123270138389e-07,
+      "loss": 0.4088,
+      "step": 1200
+    },
+    {
+      "epoch": 0.009679922560619515,
+      "grad_norm": 0.6532760834208395,
+      "learning_rate": 9.67122630189585e-07,
+      "loss": 0.3969,
+      "step": 1210
+    },
+    {
+      "epoch": 0.009759921920624635,
+      "grad_norm": 0.7563260927255472,
+      "learning_rate": 9.75121990240781e-07,
+      "loss": 0.4075,
+      "step": 1220
+    },
+    {
+      "epoch": 0.009839921280629755,
+      "grad_norm": 0.7704505024635662,
+      "learning_rate": 9.831213502919768e-07,
+      "loss": 0.3981,
+      "step": 1230
+    },
+    {
+      "epoch": 0.009919920640634875,
+      "grad_norm": 0.7056792723024738,
+      "learning_rate": 9.911207103431725e-07,
+      "loss": 0.4223,
+      "step": 1240
+    },
+    {
+      "epoch": 0.009999920000639996,
+      "grad_norm": 0.750390399783068,
+      "learning_rate": 9.991200703943684e-07,
+      "loss": 0.4102,
+      "step": 1250
+    },
+    {
+      "epoch": 0.010079919360645114,
+      "grad_norm": 0.7529460057363149,
+      "learning_rate": 1.0071194304455646e-06,
+      "loss": 0.4278,
+      "step": 1260
+    },
+    {
+      "epoch": 0.010159918720650234,
+      "grad_norm": 1.041695148553965,
+      "learning_rate": 1.0151187904967603e-06,
+      "loss": 0.4261,
+      "step": 1270
+    },
+    {
+      "epoch": 0.010239918080655355,
+      "grad_norm": 0.8098838670880232,
+      "learning_rate": 1.0231181505479562e-06,
+      "loss": 0.4163,
+      "step": 1280
+    },
+    {
+      "epoch": 0.010319917440660475,
+      "grad_norm": 0.7139180273227849,
+      "learning_rate": 1.0311175105991523e-06,
+      "loss": 0.4139,
+      "step": 1290
+    },
+    {
+      "epoch": 0.010399916800665595,
+      "grad_norm": 0.7293628900497775,
+      "learning_rate": 1.039116870650348e-06,
+      "loss": 0.4139,
+      "step": 1300
+    },
+    {
+      "epoch": 0.010479916160670715,
+      "grad_norm": 3.9899125178456276,
+      "learning_rate": 1.047116230701544e-06,
+      "loss": 0.4352,
+      "step": 1310
+    },
+    {
+      "epoch": 0.010559915520675835,
+      "grad_norm": 0.741428608083807,
+      "learning_rate": 1.0551155907527398e-06,
+      "loss": 0.4487,
+      "step": 1320
+    },
+    {
+      "epoch": 0.010639914880680954,
+      "grad_norm": 0.7645959836257802,
+      "learning_rate": 1.0631149508039358e-06,
+      "loss": 0.4509,
+      "step": 1330
+    },
+    {
+      "epoch": 0.010719914240686074,
+      "grad_norm": 0.6630163901596575,
+      "learning_rate": 1.0711143108551317e-06,
+      "loss": 0.422,
+      "step": 1340
+    },
+    {
+      "epoch": 0.010799913600691194,
+      "grad_norm": 0.7653242364061862,
+      "learning_rate": 1.0791136709063276e-06,
+      "loss": 0.4047,
+      "step": 1350
+    },
+    {
+      "epoch": 0.010879912960696314,
+      "grad_norm": 0.6952138151449674,
+      "learning_rate": 1.0871130309575235e-06,
+      "loss": 0.424,
+      "step": 1360
+    },
+    {
+      "epoch": 0.010959912320701435,
+      "grad_norm": 0.8400358998787617,
+      "learning_rate": 1.0951123910087194e-06,
+      "loss": 0.4329,
+      "step": 1370
+    },
+    {
+      "epoch": 0.011039911680706555,
+      "grad_norm": 0.7959368745719518,
+      "learning_rate": 1.1031117510599153e-06,
+      "loss": 0.4296,
+      "step": 1380
+    },
+    {
+      "epoch": 0.011119911040711675,
+      "grad_norm": 0.7182831330712102,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.4053,
+      "step": 1390
+    },
+    {
+      "epoch": 0.011199910400716794,
+      "grad_norm": 0.7631459457678664,
+      "learning_rate": 1.1191104711623072e-06,
+      "loss": 0.4174,
+      "step": 1400
+    },
+    {
+      "epoch": 0.011279909760721914,
+      "grad_norm": 0.6904060292491346,
+      "learning_rate": 1.127109831213503e-06,
+      "loss": 0.4262,
+      "step": 1410
+    },
+    {
+      "epoch": 0.011359909120727034,
+      "grad_norm": 0.7154968430224656,
+      "learning_rate": 1.1351091912646988e-06,
+      "loss": 0.4116,
+      "step": 1420
+    },
+    {
+      "epoch": 0.011439908480732154,
+      "grad_norm": 0.6430506977204669,
+      "learning_rate": 1.143108551315895e-06,
+      "loss": 0.3985,
+      "step": 1430
+    },
+    {
+      "epoch": 0.011519907840737274,
+      "grad_norm": 0.6060536937789834,
+      "learning_rate": 1.1511079113670908e-06,
+      "loss": 0.4366,
+      "step": 1440
+    },
+    {
+      "epoch": 0.011599907200742395,
+      "grad_norm": 0.7070045694797042,
+      "learning_rate": 1.1591072714182865e-06,
+      "loss": 0.3951,
+      "step": 1450
+    },
+    {
+      "epoch": 0.011679906560747515,
+      "grad_norm": 0.6799481270205386,
+      "learning_rate": 1.1671066314694824e-06,
+      "loss": 0.4277,
+      "step": 1460
+    },
+    {
+      "epoch": 0.011759905920752633,
+      "grad_norm": 0.7706688131579046,
+      "learning_rate": 1.1751059915206786e-06,
+      "loss": 0.4106,
+      "step": 1470
+    },
+    {
+      "epoch": 0.011839905280757753,
+      "grad_norm": 0.7659401655636618,
+      "learning_rate": 1.1831053515718743e-06,
+      "loss": 0.4147,
+      "step": 1480
+    },
+    {
+      "epoch": 0.011919904640762874,
+      "grad_norm": 0.7880351649260696,
+      "learning_rate": 1.1911047116230702e-06,
+      "loss": 0.4018,
+      "step": 1490
+    },
+    {
+      "epoch": 0.011999904000767994,
+      "grad_norm": 0.6643952514129879,
+      "learning_rate": 1.1991040716742661e-06,
+      "loss": 0.4225,
+      "step": 1500
+    },
+    {
+      "epoch": 0.012079903360773114,
+      "grad_norm": 0.7410891062680932,
+      "learning_rate": 1.207103431725462e-06,
+      "loss": 0.4235,
+      "step": 1510
+    },
+    {
+      "epoch": 0.012159902720778234,
+      "grad_norm": 0.6702465025094532,
+      "learning_rate": 1.215102791776658e-06,
+      "loss": 0.4485,
+      "step": 1520
+    },
+    {
+      "epoch": 0.012239902080783354,
+      "grad_norm": 0.72511805103151,
+      "learning_rate": 1.2231021518278539e-06,
+      "loss": 0.4075,
+      "step": 1530
+    },
+    {
+      "epoch": 0.012319901440788473,
+      "grad_norm": 0.7364457263640842,
+      "learning_rate": 1.2311015118790498e-06,
+      "loss": 0.4209,
+      "step": 1540
+    },
+    {
+      "epoch": 0.012399900800793593,
+      "grad_norm": 0.6760801533876919,
+      "learning_rate": 1.2391008719302457e-06,
+      "loss": 0.4153,
+      "step": 1550
+    },
+    {
+      "epoch": 0.012479900160798713,
+      "grad_norm": 1.329240774622819,
+      "learning_rate": 1.2471002319814416e-06,
+      "loss": 0.4569,
+      "step": 1560
+    },
+    {
+      "epoch": 0.012559899520803833,
+      "grad_norm": 0.7197695590369367,
+      "learning_rate": 1.2550995920326375e-06,
+      "loss": 0.4148,
+      "step": 1570
+    },
+    {
+      "epoch": 0.012639898880808954,
+      "grad_norm": 0.7197146321929337,
+      "learning_rate": 1.2630989520838332e-06,
+      "loss": 0.4309,
+      "step": 1580
+    },
+    {
+      "epoch": 0.012719898240814074,
+      "grad_norm": 0.6941046794760515,
+      "learning_rate": 1.2710983121350293e-06,
+      "loss": 0.4266,
+      "step": 1590
+    },
+    {
+      "epoch": 0.012799897600819194,
+      "grad_norm": 0.6685733935987979,
+      "learning_rate": 1.2790976721862253e-06,
+      "loss": 0.4148,
+      "step": 1600
+    },
+    {
+      "epoch": 0.012879896960824314,
+      "grad_norm": 0.829963069106683,
+      "learning_rate": 1.287097032237421e-06,
+      "loss": 0.438,
+      "step": 1610
+    },
+    {
+      "epoch": 0.012959896320829433,
+      "grad_norm": 0.8772001383565942,
+      "learning_rate": 1.295096392288617e-06,
+      "loss": 0.4167,
+      "step": 1620
+    },
+    {
+      "epoch": 0.013039895680834553,
+      "grad_norm": 0.7693529714105686,
+      "learning_rate": 1.303095752339813e-06,
+      "loss": 0.4465,
+      "step": 1630
+    },
+    {
+      "epoch": 0.013119895040839673,
+      "grad_norm": 0.6969819161826472,
+      "learning_rate": 1.3110951123910087e-06,
+      "loss": 0.4374,
+      "step": 1640
+    },
+    {
+      "epoch": 0.013199894400844793,
+      "grad_norm": 0.7245418409893126,
+      "learning_rate": 1.3190944724422048e-06,
+      "loss": 0.4375,
+      "step": 1650
+    },
+    {
+      "epoch": 0.013279893760849914,
+      "grad_norm": 0.6704658854506884,
+      "learning_rate": 1.3270938324934008e-06,
+      "loss": 0.4144,
+      "step": 1660
+    },
+    {
+      "epoch": 0.013359893120855034,
+      "grad_norm": 0.7596780468310279,
+      "learning_rate": 1.3350931925445965e-06,
+      "loss": 0.4142,
+      "step": 1670
+    },
+    {
+      "epoch": 0.013439892480860154,
+      "grad_norm": 0.799033496950235,
+      "learning_rate": 1.3430925525957924e-06,
+      "loss": 0.424,
+      "step": 1680
+    },
+    {
+      "epoch": 0.013519891840865272,
+      "grad_norm": 0.8360146268303155,
+      "learning_rate": 1.3510919126469885e-06,
+      "loss": 0.4104,
+      "step": 1690
+    },
+    {
+      "epoch": 0.013599891200870393,
+      "grad_norm": 0.7830381641790578,
+      "learning_rate": 1.3590912726981842e-06,
+      "loss": 0.4067,
+      "step": 1700
+    },
+    {
+      "epoch": 0.013679890560875513,
+      "grad_norm": 0.7376045815830132,
+      "learning_rate": 1.3670906327493801e-06,
+      "loss": 0.4232,
+      "step": 1710
+    },
+    {
+      "epoch": 0.013759889920880633,
+      "grad_norm": 0.7780257393234633,
+      "learning_rate": 1.3750899928005762e-06,
+      "loss": 0.4001,
+      "step": 1720
+    },
+    {
+      "epoch": 0.013839889280885753,
+      "grad_norm": 0.7400853428241427,
+      "learning_rate": 1.383089352851772e-06,
+      "loss": 0.4283,
+      "step": 1730
+    },
+    {
+      "epoch": 0.013919888640890873,
+      "grad_norm": 0.6301335389858812,
+      "learning_rate": 1.3910887129029679e-06,
+      "loss": 0.4315,
+      "step": 1740
+    },
+    {
+      "epoch": 0.013999888000895994,
+      "grad_norm": 0.6873058205506691,
+      "learning_rate": 1.3990880729541636e-06,
+      "loss": 0.4294,
+      "step": 1750
+    },
+    {
+      "epoch": 0.014079887360901112,
+      "grad_norm": 0.6118849450571252,
+      "learning_rate": 1.4070874330053597e-06,
+      "loss": 0.4116,
+      "step": 1760
+    },
+    {
+      "epoch": 0.014159886720906232,
+      "grad_norm": 0.8643963519063279,
+      "learning_rate": 1.4150867930565556e-06,
+      "loss": 0.4189,
+      "step": 1770
+    },
+    {
+      "epoch": 0.014239886080911352,
+      "grad_norm": 0.6986286668684888,
+      "learning_rate": 1.4230861531077513e-06,
+      "loss": 0.4393,
+      "step": 1780
+    },
+    {
+      "epoch": 0.014319885440916473,
+      "grad_norm": 0.7361554237470329,
+      "learning_rate": 1.4310855131589474e-06,
+      "loss": 0.4048,
+      "step": 1790
+    },
+    {
+      "epoch": 0.014399884800921593,
+      "grad_norm": 0.7526595512221772,
+      "learning_rate": 1.4390848732101434e-06,
+      "loss": 0.4434,
+      "step": 1800
+    },
+    {
+      "epoch": 0.014479884160926713,
+      "grad_norm": 0.6793121490590498,
+      "learning_rate": 1.447084233261339e-06,
+      "loss": 0.4313,
+      "step": 1810
+    },
+    {
+      "epoch": 0.014559883520931833,
+      "grad_norm": 0.7503593630371851,
+      "learning_rate": 1.4550835933125352e-06,
+      "loss": 0.4363,
+      "step": 1820
+    },
+    {
+      "epoch": 0.014639882880936952,
+      "grad_norm": 0.6631126747097913,
+      "learning_rate": 1.463082953363731e-06,
+      "loss": 0.4172,
+      "step": 1830
+    },
+    {
+      "epoch": 0.014719882240942072,
+      "grad_norm": 0.6201041671912539,
+      "learning_rate": 1.4710823134149268e-06,
+      "loss": 0.426,
+      "step": 1840
+    },
+    {
+      "epoch": 0.014799881600947192,
+      "grad_norm": 0.7986575840999489,
+      "learning_rate": 1.479081673466123e-06,
+      "loss": 0.4418,
+      "step": 1850
+    },
+    {
+      "epoch": 0.014879880960952312,
+      "grad_norm": 0.7485644191714067,
+      "learning_rate": 1.4870810335173188e-06,
+      "loss": 0.4315,
+      "step": 1860
+    },
+    {
+      "epoch": 0.014959880320957433,
+      "grad_norm": 0.767180980787683,
+      "learning_rate": 1.4950803935685146e-06,
+      "loss": 0.44,
+      "step": 1870
+    },
+    {
+      "epoch": 0.015039879680962553,
+      "grad_norm": 0.7085244935047633,
+      "learning_rate": 1.5030797536197107e-06,
+      "loss": 0.4197,
+      "step": 1880
+    },
+    {
+      "epoch": 0.015119879040967673,
+      "grad_norm": 0.9096558305501717,
+      "learning_rate": 1.5110791136709064e-06,
+      "loss": 0.4276,
+      "step": 1890
+    },
+    {
+      "epoch": 0.015199878400972791,
+      "grad_norm": 0.7782449144127891,
+      "learning_rate": 1.5190784737221023e-06,
+      "loss": 0.4433,
+      "step": 1900
+    },
+    {
+      "epoch": 0.015279877760977912,
+      "grad_norm": 0.6784288997728832,
+      "learning_rate": 1.5270778337732984e-06,
+      "loss": 0.408,
+      "step": 1910
+    },
+    {
+      "epoch": 0.015359877120983032,
+      "grad_norm": 0.7394160138728095,
+      "learning_rate": 1.5350771938244941e-06,
+      "loss": 0.427,
+      "step": 1920
+    },
+    {
+      "epoch": 0.015439876480988152,
+      "grad_norm": 0.7020184796096351,
+      "learning_rate": 1.54307655387569e-06,
+      "loss": 0.4585,
+      "step": 1930
+    },
+    {
+      "epoch": 0.015519875840993272,
+      "grad_norm": 0.6570216742745165,
+      "learning_rate": 1.551075913926886e-06,
+      "loss": 0.4203,
+      "step": 1940
+    },
+    {
+      "epoch": 0.015599875200998392,
+      "grad_norm": 0.6492915506893296,
+      "learning_rate": 1.5590752739780819e-06,
+      "loss": 0.4225,
+      "step": 1950
+    },
+    {
+      "epoch": 0.01567987456100351,
+      "grad_norm": 0.6404207960330748,
+      "learning_rate": 1.5670746340292778e-06,
+      "loss": 0.4155,
+      "step": 1960
+    },
+    {
+      "epoch": 0.01575987392100863,
+      "grad_norm": 0.7069937527425317,
+      "learning_rate": 1.5750739940804737e-06,
+      "loss": 0.4017,
+      "step": 1970
+    },
+    {
+      "epoch": 0.01583987328101375,
+      "grad_norm": 0.7497814910327999,
+      "learning_rate": 1.5830733541316694e-06,
+      "loss": 0.4336,
+      "step": 1980
+    },
+    {
+      "epoch": 0.01591987264101887,
+      "grad_norm": 0.997502334828043,
+      "learning_rate": 1.5910727141828655e-06,
+      "loss": 0.4252,
+      "step": 1990
+    },
+    {
+      "epoch": 0.01599987200102399,
+      "grad_norm": 0.7051737520195863,
+      "learning_rate": 1.5990720742340615e-06,
+      "loss": 0.4136,
+      "step": 2000
+    },
+    {
+      "epoch": 0.016079871361029112,
+      "grad_norm": 0.7358677233078202,
+      "learning_rate": 1.6070714342852572e-06,
+      "loss": 0.4041,
+      "step": 2010
+    },
+    {
+      "epoch": 0.016159870721034232,
+      "grad_norm": 0.7232160368639224,
+      "learning_rate": 1.6150707943364533e-06,
+      "loss": 0.3925,
+      "step": 2020
+    },
+    {
+      "epoch": 0.016239870081039352,
+      "grad_norm": 0.7492396218793221,
+      "learning_rate": 1.6230701543876492e-06,
+      "loss": 0.4453,
+      "step": 2030
+    },
+    {
+      "epoch": 0.016319869441044473,
+      "grad_norm": 0.7879468186487532,
+      "learning_rate": 1.631069514438845e-06,
+      "loss": 0.4143,
+      "step": 2040
+    },
+    {
+      "epoch": 0.016399868801049593,
+      "grad_norm": 0.7027247029038095,
+      "learning_rate": 1.639068874490041e-06,
+      "loss": 0.4487,
+      "step": 2050
+    },
+    {
+      "epoch": 0.016479868161054713,
+      "grad_norm": 0.6646702988688921,
+      "learning_rate": 1.6470682345412367e-06,
+      "loss": 0.4189,
+      "step": 2060
+    },
+    {
+      "epoch": 0.016559867521059833,
+      "grad_norm": 0.7214592394412016,
+      "learning_rate": 1.6550675945924326e-06,
+      "loss": 0.4298,
+      "step": 2070
+    },
+    {
+      "epoch": 0.01663986688106495,
+      "grad_norm": 0.7009780752105863,
+      "learning_rate": 1.6630669546436288e-06,
+      "loss": 0.43,
+      "step": 2080
+    },
+    {
+      "epoch": 0.01671986624107007,
+      "grad_norm": 0.6802031501540443,
+      "learning_rate": 1.6710663146948245e-06,
+      "loss": 0.4064,
+      "step": 2090
+    },
+    {
+      "epoch": 0.01679986560107519,
+      "grad_norm": 0.7764568933093239,
+      "learning_rate": 1.6790656747460204e-06,
+      "loss": 0.4192,
+      "step": 2100
+    },
+    {
+      "epoch": 0.01687986496108031,
+      "grad_norm": 0.7257831241745193,
+      "learning_rate": 1.6870650347972165e-06,
+      "loss": 0.4137,
+      "step": 2110
+    },
+    {
+      "epoch": 0.01695986432108543,
+      "grad_norm": 0.7231093177815116,
+      "learning_rate": 1.6950643948484122e-06,
+      "loss": 0.4321,
+      "step": 2120
+    },
+    {
+      "epoch": 0.01703986368109055,
+      "grad_norm": 0.7604848904942334,
+      "learning_rate": 1.7030637548996081e-06,
+      "loss": 0.4173,
+      "step": 2130
+    },
+    {
+      "epoch": 0.01711986304109567,
+      "grad_norm": 0.8671202797505491,
+      "learning_rate": 1.7110631149508043e-06,
+      "loss": 0.4172,
+      "step": 2140
+    },
+    {
+      "epoch": 0.01719986240110079,
+      "grad_norm": 0.8818905733108134,
+      "learning_rate": 1.719062475002e-06,
+      "loss": 0.4096,
+      "step": 2150
+    },
+    {
+      "epoch": 0.01727986176110591,
+      "grad_norm": 0.7073377083366502,
+      "learning_rate": 1.7270618350531959e-06,
+      "loss": 0.4327,
+      "step": 2160
+    },
+    {
+      "epoch": 0.01735986112111103,
+      "grad_norm": 0.9637856611728648,
+      "learning_rate": 1.735061195104392e-06,
+      "loss": 0.457,
+      "step": 2170
+    },
+    {
+      "epoch": 0.017439860481116152,
+      "grad_norm": 0.7763253238443666,
+      "learning_rate": 1.7430605551555877e-06,
+      "loss": 0.4444,
+      "step": 2180
+    },
+    {
+      "epoch": 0.017519859841121272,
+      "grad_norm": 0.7532387544984117,
+      "learning_rate": 1.7510599152067836e-06,
+      "loss": 0.4193,
+      "step": 2190
+    },
+    {
+      "epoch": 0.017599859201126392,
+      "grad_norm": 0.7001703245870058,
+      "learning_rate": 1.7590592752579793e-06,
+      "loss": 0.4334,
+      "step": 2200
+    },
+    {
+      "epoch": 0.017679858561131512,
+      "grad_norm": 0.653320909525973,
+      "learning_rate": 1.7670586353091755e-06,
+      "loss": 0.4137,
+      "step": 2210
+    },
+    {
+      "epoch": 0.01775985792113663,
+      "grad_norm": 0.6431905950303065,
+      "learning_rate": 1.7750579953603714e-06,
+      "loss": 0.4056,
+      "step": 2220
+    },
+    {
+      "epoch": 0.01783985728114175,
+      "grad_norm": 0.6355510699319944,
+      "learning_rate": 1.783057355411567e-06,
+      "loss": 0.3967,
+      "step": 2230
+    },
+    {
+      "epoch": 0.01791985664114687,
+      "grad_norm": 1.2693108108319435,
+      "learning_rate": 1.7910567154627632e-06,
+      "loss": 0.3958,
+      "step": 2240
+    },
+    {
+      "epoch": 0.01799985600115199,
+      "grad_norm": 0.7084484301635297,
+      "learning_rate": 1.7990560755139591e-06,
+      "loss": 0.4341,
+      "step": 2250
+    },
+    {
+      "epoch": 0.01807985536115711,
+      "grad_norm": 0.7305495029250423,
+      "learning_rate": 1.8070554355651548e-06,
+      "loss": 0.4343,
+      "step": 2260
+    },
+    {
+      "epoch": 0.01815985472116223,
+      "grad_norm": 0.7855576477357182,
+      "learning_rate": 1.8150547956163507e-06,
+      "loss": 0.4158,
+      "step": 2270
+    },
+    {
+      "epoch": 0.01823985408116735,
+      "grad_norm": 0.8272843321029247,
+      "learning_rate": 1.8230541556675469e-06,
+      "loss": 0.4355,
+      "step": 2280
+    },
+    {
+      "epoch": 0.01831985344117247,
+      "grad_norm": 0.753702291149743,
+      "learning_rate": 1.8310535157187426e-06,
+      "loss": 0.4025,
+      "step": 2290
+    },
+    {
+      "epoch": 0.01839985280117759,
+      "grad_norm": 0.7598104262546688,
+      "learning_rate": 1.8390528757699385e-06,
+      "loss": 0.413,
+      "step": 2300
+    },
+    {
+      "epoch": 0.01847985216118271,
+      "grad_norm": 0.7611395927729616,
+      "learning_rate": 1.8470522358211346e-06,
+      "loss": 0.4373,
+      "step": 2310
+    },
+    {
+      "epoch": 0.01855985152118783,
+      "grad_norm": 0.760722694363519,
+      "learning_rate": 1.8550515958723303e-06,
+      "loss": 0.4179,
+      "step": 2320
+    },
+    {
+      "epoch": 0.01863985088119295,
+      "grad_norm": 0.5957047498546761,
+      "learning_rate": 1.8630509559235262e-06,
+      "loss": 0.4231,
+      "step": 2330
+    },
+    {
+      "epoch": 0.01871985024119807,
+      "grad_norm": 0.7186940138009597,
+      "learning_rate": 1.8710503159747224e-06,
+      "loss": 0.4171,
+      "step": 2340
+    },
+    {
+      "epoch": 0.018799849601203192,
+      "grad_norm": 0.7760364408450707,
+      "learning_rate": 1.879049676025918e-06,
+      "loss": 0.4003,
+      "step": 2350
+    },
+    {
+      "epoch": 0.018879848961208312,
+      "grad_norm": 0.7481302705420546,
+      "learning_rate": 1.887049036077114e-06,
+      "loss": 0.4182,
+      "step": 2360
+    },
+    {
+      "epoch": 0.01895984832121343,
+      "grad_norm": 0.7358536669321117,
+      "learning_rate": 1.8950483961283097e-06,
+      "loss": 0.4272,
+      "step": 2370
+    },
+    {
+      "epoch": 0.01903984768121855,
+      "grad_norm": 0.7602582219791679,
+      "learning_rate": 1.9030477561795058e-06,
+      "loss": 0.4157,
+      "step": 2380
+    },
+    {
+      "epoch": 0.01911984704122367,
+      "grad_norm": 0.7426387012574098,
+      "learning_rate": 1.911047116230702e-06,
+      "loss": 0.4136,
+      "step": 2390
+    },
+    {
+      "epoch": 0.01919984640122879,
+      "grad_norm": 0.6954613448087369,
+      "learning_rate": 1.9190464762818974e-06,
+      "loss": 0.4081,
+      "step": 2400
+    },
+    {
+      "epoch": 0.01927984576123391,
+      "grad_norm": 0.7604294663875293,
+      "learning_rate": 1.9270458363330933e-06,
+      "loss": 0.4116,
+      "step": 2410
+    },
+    {
+      "epoch": 0.01935984512123903,
+      "grad_norm": 0.7196806209080392,
+      "learning_rate": 1.9350451963842897e-06,
+      "loss": 0.4253,
+      "step": 2420
+    },
+    {
+      "epoch": 0.01943984448124415,
+      "grad_norm": 0.7619473447002568,
+      "learning_rate": 1.943044556435485e-06,
+      "loss": 0.4048,
+      "step": 2430
+    },
+    {
+      "epoch": 0.01951984384124927,
+      "grad_norm": 0.7052222380767645,
+      "learning_rate": 1.951043916486681e-06,
+      "loss": 0.4092,
+      "step": 2440
+    },
+    {
+      "epoch": 0.01959984320125439,
+      "grad_norm": 0.6311702076235657,
+      "learning_rate": 1.9590432765378774e-06,
+      "loss": 0.409,
+      "step": 2450
+    },
+    {
+      "epoch": 0.01967984256125951,
+      "grad_norm": 0.7989650814927569,
+      "learning_rate": 1.967042636589073e-06,
+      "loss": 0.4031,
+      "step": 2460
+    },
+    {
+      "epoch": 0.01975984192126463,
+      "grad_norm": 0.7243725001067509,
+      "learning_rate": 1.975041996640269e-06,
+      "loss": 0.4139,
+      "step": 2470
+    },
+    {
+      "epoch": 0.01983984128126975,
+      "grad_norm": 0.7689695050248523,
+      "learning_rate": 1.983041356691465e-06,
+      "loss": 0.419,
+      "step": 2480
+    },
+    {
+      "epoch": 0.01991984064127487,
+      "grad_norm": 0.6784111247678405,
+      "learning_rate": 1.9910407167426607e-06,
+      "loss": 0.4255,
+      "step": 2490
+    },
+    {
+      "epoch": 0.01999984000127999,
+      "grad_norm": 0.696241896489532,
+      "learning_rate": 1.9990400767938566e-06,
+      "loss": 0.4387,
+      "step": 2500
+    },
+    {
+      "epoch": 0.020079839361285108,
+      "grad_norm": 0.7337662500397681,
+      "learning_rate": 2.0070394368450525e-06,
+      "loss": 0.4107,
+      "step": 2510
+    },
+    {
+      "epoch": 0.02015983872129023,
+      "grad_norm": 0.7522121421950386,
+      "learning_rate": 2.0150387968962484e-06,
+      "loss": 0.4341,
+      "step": 2520
+    },
+    {
+      "epoch": 0.02023983808129535,
+      "grad_norm": 0.7635117961228429,
+      "learning_rate": 2.0230381569474443e-06,
+      "loss": 0.444,
+      "step": 2530
+    },
+    {
+      "epoch": 0.02031983744130047,
+      "grad_norm": 0.7258340461733414,
+      "learning_rate": 2.0310375169986402e-06,
+      "loss": 0.4532,
+      "step": 2540
+    },
+    {
+      "epoch": 0.02039983680130559,
+      "grad_norm": 0.8227531741667117,
+      "learning_rate": 2.039036877049836e-06,
+      "loss": 0.4211,
+      "step": 2550
+    },
+    {
+      "epoch": 0.02047983616131071,
+      "grad_norm": 0.8046278992558573,
+      "learning_rate": 2.047036237101032e-06,
+      "loss": 0.4304,
+      "step": 2560
+    },
+    {
+      "epoch": 0.02055983552131583,
+      "grad_norm": 0.8583586766230429,
+      "learning_rate": 2.055035597152228e-06,
+      "loss": 0.389,
+      "step": 2570
+    },
+    {
+      "epoch": 0.02063983488132095,
+      "grad_norm": 0.7059091393994599,
+      "learning_rate": 2.063034957203424e-06,
+      "loss": 0.4143,
+      "step": 2580
+    },
+    {
+      "epoch": 0.02071983424132607,
+      "grad_norm": 0.7169812723780176,
+      "learning_rate": 2.07103431725462e-06,
+      "loss": 0.4231,
+      "step": 2590
+    },
+    {
+      "epoch": 0.02079983360133119,
+      "grad_norm": 0.744498117665369,
+      "learning_rate": 2.0790336773058157e-06,
+      "loss": 0.3931,
+      "step": 2600
+    },
+    {
+      "epoch": 0.02087983296133631,
+      "grad_norm": 0.8714679551154206,
+      "learning_rate": 2.0870330373570117e-06,
+      "loss": 0.4227,
+      "step": 2610
+    },
+    {
+      "epoch": 0.02095983232134143,
+      "grad_norm": 0.6310793333444085,
+      "learning_rate": 2.0950323974082076e-06,
+      "loss": 0.4092,
+      "step": 2620
+    },
+    {
+      "epoch": 0.02103983168134655,
+      "grad_norm": 0.7793866057944923,
+      "learning_rate": 2.1030317574594035e-06,
+      "loss": 0.4139,
+      "step": 2630
+    },
+    {
+      "epoch": 0.02111983104135167,
+      "grad_norm": 0.8117526334063988,
+      "learning_rate": 2.1110311175105994e-06,
+      "loss": 0.4154,
+      "step": 2640
+    },
+    {
+      "epoch": 0.021199830401356787,
+      "grad_norm": 0.7466901509429923,
+      "learning_rate": 2.1190304775617953e-06,
+      "loss": 0.4139,
+      "step": 2650
+    },
+    {
+      "epoch": 0.021279829761361908,
+      "grad_norm": 0.9396797684739518,
+      "learning_rate": 2.1270298376129912e-06,
+      "loss": 0.4166,
+      "step": 2660
+    },
+    {
+      "epoch": 0.021359829121367028,
+      "grad_norm": 0.7212101339600039,
+      "learning_rate": 2.135029197664187e-06,
+      "loss": 0.4299,
+      "step": 2670
+    },
+    {
+      "epoch": 0.021439828481372148,
+      "grad_norm": 0.9051397916977733,
+      "learning_rate": 2.1430285577153826e-06,
+      "loss": 0.4203,
+      "step": 2680
+    },
+    {
+      "epoch": 0.021519827841377268,
+      "grad_norm": 0.6759148258397492,
+      "learning_rate": 2.151027917766579e-06,
+      "loss": 0.4065,
+      "step": 2690
+    },
+    {
+      "epoch": 0.02159982720138239,
+      "grad_norm": 0.7191776566727898,
+      "learning_rate": 2.159027277817775e-06,
+      "loss": 0.3957,
+      "step": 2700
+    },
+    {
+      "epoch": 0.02167982656138751,
+      "grad_norm": 0.7752464373073066,
+      "learning_rate": 2.1670266378689704e-06,
+      "loss": 0.4198,
+      "step": 2710
+    },
+    {
+      "epoch": 0.02175982592139263,
+      "grad_norm": 0.7096879372714776,
+      "learning_rate": 2.1750259979201667e-06,
+      "loss": 0.4337,
+      "step": 2720
+    },
+    {
+      "epoch": 0.02183982528139775,
+      "grad_norm": 0.7577341998111574,
+      "learning_rate": 2.1830253579713626e-06,
+      "loss": 0.4484,
+      "step": 2730
+    },
+    {
+      "epoch": 0.02191982464140287,
+      "grad_norm": 0.8303467101337862,
+      "learning_rate": 2.191024718022558e-06,
+      "loss": 0.436,
+      "step": 2740
+    },
+    {
+      "epoch": 0.02199982400140799,
+      "grad_norm": 0.740434972574641,
+      "learning_rate": 2.1990240780737545e-06,
+      "loss": 0.4026,
+      "step": 2750
+    },
+    {
+      "epoch": 0.02207982336141311,
+      "grad_norm": 0.7385703007931491,
+      "learning_rate": 2.2070234381249504e-06,
+      "loss": 0.3994,
+      "step": 2760
+    },
+    {
+      "epoch": 0.02215982272141823,
+      "grad_norm": 0.804160370262748,
+      "learning_rate": 2.215022798176146e-06,
+      "loss": 0.4247,
+      "step": 2770
+    },
+    {
+      "epoch": 0.02223982208142335,
+      "grad_norm": 0.9699580286471233,
+      "learning_rate": 2.2230221582273422e-06,
+      "loss": 0.4158,
+      "step": 2780
+    },
+    {
+      "epoch": 0.02231982144142847,
+      "grad_norm": 0.854317456477185,
+      "learning_rate": 2.231021518278538e-06,
+      "loss": 0.4224,
+      "step": 2790
+    },
+    {
+      "epoch": 0.022399820801433587,
+      "grad_norm": 0.7261443031804087,
+      "learning_rate": 2.2390208783297336e-06,
+      "loss": 0.4153,
+      "step": 2800
+    },
+    {
+      "epoch": 0.022479820161438707,
+      "grad_norm": 0.7221332661288903,
+      "learning_rate": 2.24702023838093e-06,
+      "loss": 0.4287,
+      "step": 2810
+    },
+    {
+      "epoch": 0.022559819521443827,
+      "grad_norm": 0.8879662037891538,
+      "learning_rate": 2.2550195984321255e-06,
+      "loss": 0.4221,
+      "step": 2820
+    },
+    {
+      "epoch": 0.022639818881448948,
+      "grad_norm": 0.7885980884870208,
+      "learning_rate": 2.2630189584833214e-06,
+      "loss": 0.4394,
+      "step": 2830
+    },
+    {
+      "epoch": 0.022719818241454068,
+      "grad_norm": 4.841458093614334,
+      "learning_rate": 2.2710183185345173e-06,
+      "loss": 0.4306,
+      "step": 2840
+    },
+    {
+      "epoch": 0.022799817601459188,
+      "grad_norm": 0.762762416297712,
+      "learning_rate": 2.279017678585713e-06,
+      "loss": 0.4248,
+      "step": 2850
+    },
+    {
+      "epoch": 0.022879816961464308,
+      "grad_norm": 0.8448845187481899,
+      "learning_rate": 2.287017038636909e-06,
+      "loss": 0.4016,
+      "step": 2860
+    },
+    {
+      "epoch": 0.02295981632146943,
+      "grad_norm": 0.7119692504285716,
+      "learning_rate": 2.295016398688105e-06,
+      "loss": 0.4187,
+      "step": 2870
+    },
+    {
+      "epoch": 0.02303981568147455,
+      "grad_norm": 0.7322860524669678,
+      "learning_rate": 2.303015758739301e-06,
+      "loss": 0.3992,
+      "step": 2880
+    },
+    {
+      "epoch": 0.02311981504147967,
+      "grad_norm": 0.7809553100217358,
+      "learning_rate": 2.311015118790497e-06,
+      "loss": 0.4323,
+      "step": 2890
+    },
+    {
+      "epoch": 0.02319981440148479,
+      "grad_norm": 0.7497042609672665,
+      "learning_rate": 2.3190144788416928e-06,
+      "loss": 0.4216,
+      "step": 2900
+    },
+    {
+      "epoch": 0.02327981376148991,
+      "grad_norm": 0.77908486703842,
+      "learning_rate": 2.3270138388928887e-06,
+      "loss": 0.4139,
+      "step": 2910
+    },
+    {
+      "epoch": 0.02335981312149503,
+      "grad_norm": 0.8259941131428074,
+      "learning_rate": 2.3350131989440846e-06,
+      "loss": 0.4298,
+      "step": 2920
+    },
+    {
+      "epoch": 0.02343981248150015,
+      "grad_norm": 0.7315958297934407,
+      "learning_rate": 2.3430125589952805e-06,
+      "loss": 0.4148,
+      "step": 2930
+    },
+    {
+      "epoch": 0.023519811841505266,
+      "grad_norm": 0.7127501348213877,
+      "learning_rate": 2.3510119190464764e-06,
+      "loss": 0.4226,
+      "step": 2940
+    },
+    {
+      "epoch": 0.023599811201510387,
+      "grad_norm": 0.7004822200975431,
+      "learning_rate": 2.3590112790976724e-06,
+      "loss": 0.4037,
+      "step": 2950
+    },
+    {
+      "epoch": 0.023679810561515507,
+      "grad_norm": 0.7640873416448367,
+      "learning_rate": 2.3670106391488683e-06,
+      "loss": 0.4196,
+      "step": 2960
+    },
+    {
+      "epoch": 0.023759809921520627,
+      "grad_norm": 0.7562584220469137,
+      "learning_rate": 2.375009999200064e-06,
+      "loss": 0.4226,
+      "step": 2970
+    },
+    {
+      "epoch": 0.023839809281525747,
+      "grad_norm": 0.700408619087647,
+      "learning_rate": 2.38300935925126e-06,
+      "loss": 0.4194,
+      "step": 2980
+    },
+    {
+      "epoch": 0.023919808641530867,
+      "grad_norm": 0.6926553567290514,
+      "learning_rate": 2.391008719302456e-06,
+      "loss": 0.4377,
+      "step": 2990
+    },
+    {
+      "epoch": 0.023999808001535988,
+      "grad_norm": 0.8080569171783707,
+      "learning_rate": 2.399008079353652e-06,
+      "loss": 0.425,
+      "step": 3000
+    },
+    {
+      "epoch": 0.024079807361541108,
+      "grad_norm": 0.7127303977340983,
+      "learning_rate": 2.407007439404848e-06,
+      "loss": 0.4132,
+      "step": 3010
+    },
+    {
+      "epoch": 0.024159806721546228,
+      "grad_norm": 0.8197933173831502,
+      "learning_rate": 2.4150067994560438e-06,
+      "loss": 0.455,
+      "step": 3020
+    },
+    {
+      "epoch": 0.024239806081551348,
+      "grad_norm": 0.7710243766351836,
+      "learning_rate": 2.4230061595072397e-06,
+      "loss": 0.4181,
+      "step": 3030
+    },
+    {
+      "epoch": 0.02431980544155647,
+      "grad_norm": 0.7237467513596426,
+      "learning_rate": 2.4310055195584356e-06,
+      "loss": 0.4019,
+      "step": 3040
+    },
+    {
+      "epoch": 0.02439980480156159,
+      "grad_norm": 0.6859301673181247,
+      "learning_rate": 2.4390048796096315e-06,
+      "loss": 0.417,
+      "step": 3050
+    },
+    {
+      "epoch": 0.02447980416156671,
+      "grad_norm": 0.7468441788531288,
+      "learning_rate": 2.4470042396608274e-06,
+      "loss": 0.4332,
+      "step": 3060
+    },
+    {
+      "epoch": 0.02455980352157183,
+      "grad_norm": 0.6969645352923937,
+      "learning_rate": 2.4550035997120233e-06,
+      "loss": 0.4253,
+      "step": 3070
+    },
+    {
+      "epoch": 0.024639802881576946,
+      "grad_norm": 0.732458270879925,
+      "learning_rate": 2.4630029597632193e-06,
+      "loss": 0.4071,
+      "step": 3080
+    },
+    {
+      "epoch": 0.024719802241582066,
+      "grad_norm": 1.0213686642017104,
+      "learning_rate": 2.471002319814415e-06,
+      "loss": 0.4284,
+      "step": 3090
+    },
+    {
+      "epoch": 0.024799801601587186,
+      "grad_norm": 0.7109617020682577,
+      "learning_rate": 2.479001679865611e-06,
+      "loss": 0.4338,
+      "step": 3100
+    },
+    {
+      "epoch": 0.024879800961592306,
+      "grad_norm": 0.7970001881138196,
+      "learning_rate": 2.487001039916807e-06,
+      "loss": 0.4071,
+      "step": 3110
+    },
+    {
+      "epoch": 0.024959800321597427,
+      "grad_norm": 0.7531634719565369,
+      "learning_rate": 2.495000399968003e-06,
+      "loss": 0.4639,
+      "step": 3120
+    },
+    {
+      "epoch": 0.025039799681602547,
+      "grad_norm": 0.6656181891345934,
+      "learning_rate": 2.5029997600191984e-06,
+      "loss": 0.4099,
+      "step": 3130
+    },
+    {
+      "epoch": 0.025119799041607667,
+      "grad_norm": 0.9053194365491155,
+      "learning_rate": 2.5109991200703947e-06,
+      "loss": 0.4117,
+      "step": 3140
+    },
+    {
+      "epoch": 0.025199798401612787,
+      "grad_norm": 0.7986025391471563,
+      "learning_rate": 2.5189984801215907e-06,
+      "loss": 0.4293,
+      "step": 3150
+    },
+    {
+      "epoch": 0.025279797761617907,
+      "grad_norm": 0.7216163241881903,
+      "learning_rate": 2.526997840172786e-06,
+      "loss": 0.4014,
+      "step": 3160
+    },
+    {
+      "epoch": 0.025359797121623028,
+      "grad_norm": 0.8773953768019375,
+      "learning_rate": 2.5349972002239825e-06,
+      "loss": 0.4272,
+      "step": 3170
+    },
+    {
+      "epoch": 0.025439796481628148,
+      "grad_norm": 0.69671419994169,
+      "learning_rate": 2.5429965602751784e-06,
+      "loss": 0.4056,
+      "step": 3180
+    },
+    {
+      "epoch": 0.025519795841633268,
+      "grad_norm": 0.8372255728456501,
+      "learning_rate": 2.550995920326374e-06,
+      "loss": 0.421,
+      "step": 3190
+    },
+    {
+      "epoch": 0.025599795201638388,
+      "grad_norm": 0.7425111234564166,
+      "learning_rate": 2.5589952803775702e-06,
+      "loss": 0.42,
+      "step": 3200
+    },
+    {
+      "epoch": 0.02567979456164351,
+      "grad_norm": 0.7792291619496273,
+      "learning_rate": 2.566994640428766e-06,
+      "loss": 0.3956,
+      "step": 3210
+    },
+    {
+      "epoch": 0.02575979392164863,
+      "grad_norm": 0.791207552625433,
+      "learning_rate": 2.5749940004799616e-06,
+      "loss": 0.4284,
+      "step": 3220
+    },
+    {
+      "epoch": 0.025839793281653745,
+      "grad_norm": 0.7836824912313216,
+      "learning_rate": 2.5829933605311576e-06,
+      "loss": 0.4091,
+      "step": 3230
+    },
+    {
+      "epoch": 0.025919792641658865,
+      "grad_norm": 0.7110649040200595,
+      "learning_rate": 2.590992720582354e-06,
+      "loss": 0.4135,
+      "step": 3240
+    },
+    {
+      "epoch": 0.025999792001663986,
+      "grad_norm": 0.7404153916766049,
+      "learning_rate": 2.5989920806335494e-06,
+      "loss": 0.429,
+      "step": 3250
+    },
+    {
+      "epoch": 0.026079791361669106,
+      "grad_norm": 0.6986558695921249,
+      "learning_rate": 2.6069914406847453e-06,
+      "loss": 0.4334,
+      "step": 3260
+    },
+    {
+      "epoch": 0.026159790721674226,
+      "grad_norm": 0.7632258973881477,
+      "learning_rate": 2.6149908007359416e-06,
+      "loss": 0.3944,
+      "step": 3270
+    },
+    {
+      "epoch": 0.026239790081679346,
+      "grad_norm": 0.785186697637848,
+      "learning_rate": 2.622990160787137e-06,
+      "loss": 0.4178,
+      "step": 3280
+    },
+    {
+      "epoch": 0.026319789441684466,
+      "grad_norm": 0.8661439723668041,
+      "learning_rate": 2.630989520838333e-06,
+      "loss": 0.4113,
+      "step": 3290
+    },
+    {
+      "epoch": 0.026399788801689587,
+      "grad_norm": 0.7237402851018382,
+      "learning_rate": 2.6389888808895294e-06,
+      "loss": 0.4078,
+      "step": 3300
+    },
+    {
+      "epoch": 0.026479788161694707,
+      "grad_norm": 0.8392950022908678,
+      "learning_rate": 2.646988240940725e-06,
+      "loss": 0.4223,
+      "step": 3310
+    },
+    {
+      "epoch": 0.026559787521699827,
+      "grad_norm": 0.92567701672041,
+      "learning_rate": 2.654987600991921e-06,
+      "loss": 0.3845,
+      "step": 3320
+    },
+    {
+      "epoch": 0.026639786881704947,
+      "grad_norm": 0.6933742034635484,
+      "learning_rate": 2.662986961043117e-06,
+      "loss": 0.427,
+      "step": 3330
+    },
+    {
+      "epoch": 0.026719786241710067,
+      "grad_norm": 0.7620381540647038,
+      "learning_rate": 2.6709863210943126e-06,
+      "loss": 0.4209,
+      "step": 3340
+    },
+    {
+      "epoch": 0.026799785601715188,
+      "grad_norm": 0.721902719522637,
+      "learning_rate": 2.6789856811455085e-06,
+      "loss": 0.41,
+      "step": 3350
+    },
+    {
+      "epoch": 0.026879784961720308,
+      "grad_norm": 0.7299501620253848,
+      "learning_rate": 2.6869850411967045e-06,
+      "loss": 0.4166,
+      "step": 3360
+    },
+    {
+      "epoch": 0.026959784321725425,
+      "grad_norm": 0.7715065662581788,
+      "learning_rate": 2.6949844012479004e-06,
+      "loss": 0.4024,
+      "step": 3370
+    },
+    {
+      "epoch": 0.027039783681730545,
+      "grad_norm": 0.8682048882317643,
+      "learning_rate": 2.7029837612990963e-06,
+      "loss": 0.4131,
+      "step": 3380
+    },
+    {
+      "epoch": 0.027119783041735665,
+      "grad_norm": 0.869764700266281,
+      "learning_rate": 2.710983121350292e-06,
+      "loss": 0.4261,
+      "step": 3390
+    },
+    {
+      "epoch": 0.027199782401740785,
+      "grad_norm": 0.7340237050441278,
+      "learning_rate": 2.718982481401488e-06,
+      "loss": 0.4025,
+      "step": 3400
+    },
+    {
+      "epoch": 0.027279781761745905,
+      "grad_norm": 0.8579554974753395,
+      "learning_rate": 2.726981841452684e-06,
+      "loss": 0.407,
+      "step": 3410
+    },
+    {
+      "epoch": 0.027359781121751026,
+      "grad_norm": 0.7274223844530919,
+      "learning_rate": 2.73498120150388e-06,
+      "loss": 0.4133,
+      "step": 3420
+    },
+    {
+      "epoch": 0.027439780481756146,
+      "grad_norm": 0.7441058560749837,
+      "learning_rate": 2.742980561555076e-06,
+      "loss": 0.4243,
+      "step": 3430
+    },
+    {
+      "epoch": 0.027519779841761266,
+      "grad_norm": 0.8051355651618886,
+      "learning_rate": 2.7509799216062718e-06,
+      "loss": 0.4251,
+      "step": 3440
+    },
+    {
+      "epoch": 0.027599779201766386,
+      "grad_norm": 0.7432285405318118,
+      "learning_rate": 2.7589792816574677e-06,
+      "loss": 0.4412,
+      "step": 3450
+    },
+    {
+      "epoch": 0.027679778561771506,
+      "grad_norm": 0.6517122251801726,
+      "learning_rate": 2.7669786417086636e-06,
+      "loss": 0.4252,
+      "step": 3460
+    },
+    {
+      "epoch": 0.027759777921776627,
+      "grad_norm": 0.6875721581221387,
+      "learning_rate": 2.774978001759859e-06,
+      "loss": 0.4177,
+      "step": 3470
+    },
+    {
+      "epoch": 0.027839777281781747,
+      "grad_norm": 0.702253196059382,
+      "learning_rate": 2.7829773618110554e-06,
+      "loss": 0.4205,
+      "step": 3480
+    },
+    {
+      "epoch": 0.027919776641786867,
+      "grad_norm": 0.8482036395209255,
+      "learning_rate": 2.7909767218622514e-06,
+      "loss": 0.4017,
+      "step": 3490
+    },
+    {
+      "epoch": 0.027999776001791987,
+      "grad_norm": 0.6925655746894002,
+      "learning_rate": 2.798976081913447e-06,
+      "loss": 0.4121,
+      "step": 3500
+    },
+    {
+      "epoch": 0.028079775361797104,
+      "grad_norm": 0.7220552449627191,
+      "learning_rate": 2.806975441964643e-06,
+      "loss": 0.4102,
+      "step": 3510
+    },
+    {
+      "epoch": 0.028159774721802224,
+      "grad_norm": 0.8831384842937139,
+      "learning_rate": 2.814974802015839e-06,
+      "loss": 0.4228,
+      "step": 3520
+    },
+    {
+      "epoch": 0.028239774081807344,
+      "grad_norm": 0.7579479724344111,
+      "learning_rate": 2.8229741620670346e-06,
+      "loss": 0.413,
+      "step": 3530
+    },
+    {
+      "epoch": 0.028319773441812465,
+      "grad_norm": 0.7501882165699492,
+      "learning_rate": 2.8309735221182305e-06,
+      "loss": 0.4282,
+      "step": 3540
+    },
+    {
+      "epoch": 0.028399772801817585,
+      "grad_norm": 0.8755928070904447,
+      "learning_rate": 2.838972882169427e-06,
+      "loss": 0.4007,
+      "step": 3550
+    },
+    {
+      "epoch": 0.028479772161822705,
+      "grad_norm": 0.7554891461392241,
+      "learning_rate": 2.8469722422206223e-06,
+      "loss": 0.4168,
+      "step": 3560
+    },
+    {
+      "epoch": 0.028559771521827825,
+      "grad_norm": 0.6985247419196446,
+      "learning_rate": 2.8549716022718183e-06,
+      "loss": 0.4169,
+      "step": 3570
+    },
+    {
+      "epoch": 0.028639770881832945,
+      "grad_norm": 0.7592211529186911,
+      "learning_rate": 2.8629709623230146e-06,
+      "loss": 0.4065,
+      "step": 3580
+    },
+    {
+      "epoch": 0.028719770241838066,
+      "grad_norm": 0.7833965541238131,
+      "learning_rate": 2.87097032237421e-06,
+      "loss": 0.4446,
+      "step": 3590
+    },
+    {
+      "epoch": 0.028799769601843186,
+      "grad_norm": 0.7738713593579586,
+      "learning_rate": 2.878969682425406e-06,
+      "loss": 0.419,
+      "step": 3600
+    },
+    {
+      "epoch": 0.028879768961848306,
+      "grad_norm": 0.8315629244025291,
+      "learning_rate": 2.8869690424766023e-06,
+      "loss": 0.4219,
+      "step": 3610
+    },
+    {
+      "epoch": 0.028959768321853426,
+      "grad_norm": 0.781696668298691,
+      "learning_rate": 2.894968402527798e-06,
+      "loss": 0.4151,
+      "step": 3620
+    },
+    {
+      "epoch": 0.029039767681858546,
+      "grad_norm": 0.7720539381394058,
+      "learning_rate": 2.9029677625789938e-06,
+      "loss": 0.4202,
+      "step": 3630
+    },
+    {
+      "epoch": 0.029119767041863667,
+      "grad_norm": 0.7526278482938622,
+      "learning_rate": 2.91096712263019e-06,
+      "loss": 0.4436,
+      "step": 3640
+    },
+    {
+      "epoch": 0.029199766401868787,
+      "grad_norm": 0.7337976104344518,
+      "learning_rate": 2.9189664826813856e-06,
+      "loss": 0.4236,
+      "step": 3650
+    },
+    {
+      "epoch": 0.029279765761873904,
+      "grad_norm": 0.83533803326155,
+      "learning_rate": 2.9269658427325815e-06,
+      "loss": 0.4176,
+      "step": 3660
+    },
+    {
+      "epoch": 0.029359765121879024,
+      "grad_norm": 0.787318610210915,
+      "learning_rate": 2.934965202783778e-06,
+      "loss": 0.4165,
+      "step": 3670
+    },
+    {
+      "epoch": 0.029439764481884144,
+      "grad_norm": 1.4090517215188612,
+      "learning_rate": 2.9429645628349733e-06,
+      "loss": 0.4119,
+      "step": 3680
+    },
+    {
+      "epoch": 0.029519763841889264,
+      "grad_norm": 0.9266366914858271,
+      "learning_rate": 2.9509639228861692e-06,
+      "loss": 0.4179,
+      "step": 3690
+    },
+    {
+      "epoch": 0.029599763201894384,
+      "grad_norm": 0.934424674393174,
+      "learning_rate": 2.9589632829373656e-06,
+      "loss": 0.4,
+      "step": 3700
+    },
+    {
+      "epoch": 0.029679762561899505,
+      "grad_norm": 0.8568902144006225,
+      "learning_rate": 2.966962642988561e-06,
+      "loss": 0.4096,
+      "step": 3710
+    },
+    {
+      "epoch": 0.029759761921904625,
+      "grad_norm": 0.8114058465400533,
+      "learning_rate": 2.974962003039757e-06,
+      "loss": 0.4447,
+      "step": 3720
+    },
+    {
+      "epoch": 0.029839761281909745,
+      "grad_norm": 1.6052989805278164,
+      "learning_rate": 2.9829613630909533e-06,
+      "loss": 0.4291,
+      "step": 3730
+    },
+    {
+      "epoch": 0.029919760641914865,
+      "grad_norm": 0.73538618969903,
+      "learning_rate": 2.990960723142149e-06,
+      "loss": 0.4248,
+      "step": 3740
+    },
+    {
+      "epoch": 0.029999760001919985,
+      "grad_norm": 0.7966165280123241,
+      "learning_rate": 2.9989600831933447e-06,
+      "loss": 0.4153,
+      "step": 3750
+    },
+    {
+      "epoch": 0.030079759361925106,
+      "grad_norm": 0.8380095957922997,
+      "learning_rate": 3.006959443244541e-06,
+      "loss": 0.378,
+      "step": 3760
+    },
+    {
+      "epoch": 0.030159758721930226,
+      "grad_norm": 0.9346417813995904,
+      "learning_rate": 3.0149588032957366e-06,
+      "loss": 0.4242,
+      "step": 3770
+    },
+    {
+      "epoch": 0.030239758081935346,
+      "grad_norm": 0.7661127998339704,
+      "learning_rate": 3.0229581633469325e-06,
+      "loss": 0.4218,
+      "step": 3780
+    },
+    {
+      "epoch": 0.030319757441940466,
+      "grad_norm": 0.7820076796497493,
+      "learning_rate": 3.030957523398129e-06,
+      "loss": 0.4057,
+      "step": 3790
+    },
+    {
+      "epoch": 0.030399756801945583,
+      "grad_norm": 0.7683193002692328,
+      "learning_rate": 3.0389568834493243e-06,
+      "loss": 0.4282,
+      "step": 3800
+    },
+    {
+      "epoch": 0.030479756161950703,
+      "grad_norm": 0.7330718238871217,
+      "learning_rate": 3.0469562435005202e-06,
+      "loss": 0.4155,
+      "step": 3810
+    },
+    {
+      "epoch": 0.030559755521955823,
+      "grad_norm": 0.7070524082806218,
+      "learning_rate": 3.0549556035517166e-06,
+      "loss": 0.4087,
+      "step": 3820
+    },
+    {
+      "epoch": 0.030639754881960943,
+      "grad_norm": 0.7825560701966258,
+      "learning_rate": 3.062954963602912e-06,
+      "loss": 0.4149,
+      "step": 3830
+    },
+    {
+      "epoch": 0.030719754241966064,
+      "grad_norm": 0.674902754654649,
+      "learning_rate": 3.070954323654108e-06,
+      "loss": 0.4222,
+      "step": 3840
+    },
+    {
+      "epoch": 0.030799753601971184,
+      "grad_norm": 0.6899663241225518,
+      "learning_rate": 3.0789536837053035e-06,
+      "loss": 0.4072,
+      "step": 3850
+    },
+    {
+      "epoch": 0.030879752961976304,
+      "grad_norm": 0.747751536532187,
+      "learning_rate": 3.0869530437565e-06,
+      "loss": 0.4265,
+      "step": 3860
+    },
+    {
+      "epoch": 0.030959752321981424,
+      "grad_norm": 0.7319670446839773,
+      "learning_rate": 3.0949524038076957e-06,
+      "loss": 0.4365,
+      "step": 3870
+    },
+    {
+      "epoch": 0.031039751681986544,
+      "grad_norm": 0.8044391132448001,
+      "learning_rate": 3.1029517638588912e-06,
+      "loss": 0.4345,
+      "step": 3880
+    },
+    {
+      "epoch": 0.031119751041991665,
+      "grad_norm": 0.6852016073842015,
+      "learning_rate": 3.1109511239100876e-06,
+      "loss": 0.3909,
+      "step": 3890
+    },
+    {
+      "epoch": 0.031199750401996785,
+      "grad_norm": 0.8109666979474524,
+      "learning_rate": 3.1189504839612835e-06,
+      "loss": 0.4159,
+      "step": 3900
+    },
+    {
+      "epoch": 0.0312797497620019,
+      "grad_norm": 0.7690716259319208,
+      "learning_rate": 3.126949844012479e-06,
+      "loss": 0.4261,
+      "step": 3910
+    },
+    {
+      "epoch": 0.03135974912200702,
+      "grad_norm": 0.770382066563297,
+      "learning_rate": 3.1349492040636753e-06,
+      "loss": 0.4197,
+      "step": 3920
+    },
+    {
+      "epoch": 0.03143974848201214,
+      "grad_norm": 0.7899366284185928,
+      "learning_rate": 3.1429485641148712e-06,
+      "loss": 0.4356,
+      "step": 3930
+    },
+    {
+      "epoch": 0.03151974784201726,
+      "grad_norm": 0.7287030959080197,
+      "learning_rate": 3.1509479241660667e-06,
+      "loss": 0.4271,
+      "step": 3940
+    },
+    {
+      "epoch": 0.03159974720202238,
+      "grad_norm": 0.6751549357858156,
+      "learning_rate": 3.158947284217263e-06,
+      "loss": 0.432,
+      "step": 3950
+    },
+    {
+      "epoch": 0.0316797465620275,
+      "grad_norm": 0.7162873288902682,
+      "learning_rate": 3.166946644268459e-06,
+      "loss": 0.4168,
+      "step": 3960
+    },
+    {
+      "epoch": 0.03175974592203262,
+      "grad_norm": 0.7624399562051731,
+      "learning_rate": 3.1749460043196545e-06,
+      "loss": 0.4077,
+      "step": 3970
+    },
+    {
+      "epoch": 0.03183974528203774,
+      "grad_norm": 0.720382439048551,
+      "learning_rate": 3.182945364370851e-06,
+      "loss": 0.4186,
+      "step": 3980
+    },
+    {
+      "epoch": 0.03191974464204286,
+      "grad_norm": 0.8771157238336784,
+      "learning_rate": 3.1909447244220467e-06,
+      "loss": 0.411,
+      "step": 3990
+    },
+    {
+      "epoch": 0.03199974400204798,
+      "grad_norm": 0.8439843188095638,
+      "learning_rate": 3.198944084473242e-06,
+      "loss": 0.4254,
+      "step": 4000
+    },
+    {
+      "epoch": 0.032079743362053104,
+      "grad_norm": 0.7522465189570665,
+      "learning_rate": 3.2069434445244385e-06,
+      "loss": 0.4182,
+      "step": 4010
+    },
+    {
+      "epoch": 0.032159742722058224,
+      "grad_norm": 0.7990809626930068,
+      "learning_rate": 3.214942804575634e-06,
+      "loss": 0.4283,
+      "step": 4020
+    },
+    {
+      "epoch": 0.032239742082063344,
+      "grad_norm": 0.7484878445246484,
+      "learning_rate": 3.22294216462683e-06,
+      "loss": 0.3957,
+      "step": 4030
+    },
+    {
+      "epoch": 0.032319741442068464,
+      "grad_norm": 0.653062676131544,
+      "learning_rate": 3.2309415246780263e-06,
+      "loss": 0.4339,
+      "step": 4040
+    },
+    {
+      "epoch": 0.032399740802073584,
+      "grad_norm": 0.7654847826834672,
+      "learning_rate": 3.2389408847292218e-06,
+      "loss": 0.4277,
+      "step": 4050
+    },
+    {
+      "epoch": 0.032479740162078705,
+      "grad_norm": 0.7545901814731136,
+      "learning_rate": 3.2469402447804177e-06,
+      "loss": 0.4181,
+      "step": 4060
+    },
+    {
+      "epoch": 0.032559739522083825,
+      "grad_norm": 0.6639250152597482,
+      "learning_rate": 3.254939604831614e-06,
+      "loss": 0.4259,
+      "step": 4070
+    },
+    {
+      "epoch": 0.032639738882088945,
+      "grad_norm": 0.8320704412806774,
+      "learning_rate": 3.2629389648828095e-06,
+      "loss": 0.4258,
+      "step": 4080
+    },
+    {
+      "epoch": 0.032719738242094065,
+      "grad_norm": 0.8748220246905324,
+      "learning_rate": 3.2709383249340054e-06,
+      "loss": 0.4348,
+      "step": 4090
+    },
+    {
+      "epoch": 0.032799737602099185,
+      "grad_norm": 1.0006726922931282,
+      "learning_rate": 3.2789376849852018e-06,
+      "loss": 0.4111,
+      "step": 4100
+    },
+    {
+      "epoch": 0.032879736962104306,
+      "grad_norm": 0.703908274983939,
+      "learning_rate": 3.2869370450363973e-06,
+      "loss": 0.4508,
+      "step": 4110
+    },
+    {
+      "epoch": 0.032959736322109426,
+      "grad_norm": 0.7114379493698554,
+      "learning_rate": 3.294936405087593e-06,
+      "loss": 0.4011,
+      "step": 4120
+    },
+    {
+      "epoch": 0.033039735682114546,
+      "grad_norm": 0.7836668838529308,
+      "learning_rate": 3.3029357651387895e-06,
+      "loss": 0.4059,
+      "step": 4130
+    },
+    {
+      "epoch": 0.033119735042119666,
+      "grad_norm": 0.7037822230916331,
+      "learning_rate": 3.310935125189985e-06,
+      "loss": 0.4171,
+      "step": 4140
+    },
+    {
+      "epoch": 0.033199734402124786,
+      "grad_norm": 0.8429288510951353,
+      "learning_rate": 3.318934485241181e-06,
+      "loss": 0.4336,
+      "step": 4150
+    },
+    {
+      "epoch": 0.0332797337621299,
+      "grad_norm": 0.8160186312241915,
+      "learning_rate": 3.3269338452923764e-06,
+      "loss": 0.4313,
+      "step": 4160
+    },
+    {
+      "epoch": 0.03335973312213502,
+      "grad_norm": 0.8158956823402862,
+      "learning_rate": 3.3349332053435728e-06,
+      "loss": 0.4117,
+      "step": 4170
+    },
+    {
+      "epoch": 0.03343973248214014,
+      "grad_norm": 0.7830740452600919,
+      "learning_rate": 3.3429325653947687e-06,
+      "loss": 0.394,
+      "step": 4180
+    },
+    {
+      "epoch": 0.03351973184214526,
+      "grad_norm": 0.8744626314539858,
+      "learning_rate": 3.350931925445964e-06,
+      "loss": 0.4229,
+      "step": 4190
+    },
+    {
+      "epoch": 0.03359973120215038,
+      "grad_norm": 0.8927610578944701,
+      "learning_rate": 3.3589312854971605e-06,
+      "loss": 0.4582,
+      "step": 4200
+    },
+    {
+      "epoch": 0.0336797305621555,
+      "grad_norm": 0.7195985669161646,
+      "learning_rate": 3.3669306455483564e-06,
+      "loss": 0.4335,
+      "step": 4210
+    },
+    {
+      "epoch": 0.03375972992216062,
+      "grad_norm": 0.8906544223501964,
+      "learning_rate": 3.374930005599552e-06,
+      "loss": 0.4131,
+      "step": 4220
+    },
+    {
+      "epoch": 0.03383972928216574,
+      "grad_norm": 0.7389006577317037,
+      "learning_rate": 3.3829293656507483e-06,
+      "loss": 0.4272,
+      "step": 4230
+    },
+    {
+      "epoch": 0.03391972864217086,
+      "grad_norm": 0.742682266765258,
+      "learning_rate": 3.390928725701944e-06,
+      "loss": 0.4418,
+      "step": 4240
+    },
+    {
+      "epoch": 0.03399972800217598,
+      "grad_norm": 0.6774990098774326,
+      "learning_rate": 3.3989280857531397e-06,
+      "loss": 0.4194,
+      "step": 4250
+    },
+    {
+      "epoch": 0.0340797273621811,
+      "grad_norm": 0.9017975115557092,
+      "learning_rate": 3.406927445804336e-06,
+      "loss": 0.4148,
+      "step": 4260
+    },
+    {
+      "epoch": 0.03415972672218622,
+      "grad_norm": 0.7641682976170854,
+      "learning_rate": 3.414926805855532e-06,
+      "loss": 0.4015,
+      "step": 4270
+    },
+    {
+      "epoch": 0.03423972608219134,
+      "grad_norm": 1.7976847080628966,
+      "learning_rate": 3.4229261659067274e-06,
+      "loss": 0.4218,
+      "step": 4280
+    },
+    {
+      "epoch": 0.03431972544219646,
+      "grad_norm": 0.7912987344925629,
+      "learning_rate": 3.4309255259579237e-06,
+      "loss": 0.4273,
+      "step": 4290
+    },
+    {
+      "epoch": 0.03439972480220158,
+      "grad_norm": 0.865219190435078,
+      "learning_rate": 3.4389248860091197e-06,
+      "loss": 0.4155,
+      "step": 4300
+    },
+    {
+      "epoch": 0.0344797241622067,
+      "grad_norm": 0.7247465267197502,
+      "learning_rate": 3.446924246060315e-06,
+      "loss": 0.4354,
+      "step": 4310
+    },
+    {
+      "epoch": 0.03455972352221182,
+      "grad_norm": 0.7302003644690978,
+      "learning_rate": 3.4549236061115115e-06,
+      "loss": 0.4202,
+      "step": 4320
+    },
+    {
+      "epoch": 0.03463972288221694,
+      "grad_norm": 0.7687470691197218,
+      "learning_rate": 3.4629229661627074e-06,
+      "loss": 0.4162,
+      "step": 4330
+    },
+    {
+      "epoch": 0.03471972224222206,
+      "grad_norm": 0.8026747308203839,
+      "learning_rate": 3.470922326213903e-06,
+      "loss": 0.4168,
+      "step": 4340
+    },
+    {
+      "epoch": 0.034799721602227184,
+      "grad_norm": 0.7280393352379876,
+      "learning_rate": 3.4789216862650992e-06,
+      "loss": 0.4196,
+      "step": 4350
+    },
+    {
+      "epoch": 0.034879720962232304,
+      "grad_norm": 0.7564739490939029,
+      "learning_rate": 3.486921046316295e-06,
+      "loss": 0.4385,
+      "step": 4360
+    },
+    {
+      "epoch": 0.034959720322237424,
+      "grad_norm": 0.8650076738362572,
+      "learning_rate": 3.4949204063674906e-06,
+      "loss": 0.4092,
+      "step": 4370
+    },
+    {
+      "epoch": 0.035039719682242544,
+      "grad_norm": 0.7393962081468373,
+      "learning_rate": 3.502919766418687e-06,
+      "loss": 0.424,
+      "step": 4380
+    },
+    {
+      "epoch": 0.035119719042247664,
+      "grad_norm": 0.9059758990047778,
+      "learning_rate": 3.510919126469883e-06,
+      "loss": 0.4222,
+      "step": 4390
+    },
+    {
+      "epoch": 0.035199718402252785,
+      "grad_norm": 0.7241080903311046,
+      "learning_rate": 3.5189184865210784e-06,
+      "loss": 0.4176,
+      "step": 4400
+    },
+    {
+      "epoch": 0.035279717762257905,
+      "grad_norm": 0.7393352165228971,
+      "learning_rate": 3.5269178465722747e-06,
+      "loss": 0.4193,
+      "step": 4410
+    },
+    {
+      "epoch": 0.035359717122263025,
+      "grad_norm": 0.7178879442718987,
+      "learning_rate": 3.5349172066234706e-06,
+      "loss": 0.4219,
+      "step": 4420
+    },
+    {
+      "epoch": 0.035439716482268145,
+      "grad_norm": 0.8075219805081396,
+      "learning_rate": 3.542916566674666e-06,
+      "loss": 0.4147,
+      "step": 4430
+    },
+    {
+      "epoch": 0.03551971584227326,
+      "grad_norm": 1.2127017079493791,
+      "learning_rate": 3.5509159267258625e-06,
+      "loss": 0.4132,
+      "step": 4440
+    },
+    {
+      "epoch": 0.03559971520227838,
+      "grad_norm": 0.7192010045742855,
+      "learning_rate": 3.5589152867770584e-06,
+      "loss": 0.411,
+      "step": 4450
+    },
+    {
+      "epoch": 0.0356797145622835,
+      "grad_norm": 0.9428723692964156,
+      "learning_rate": 3.566914646828254e-06,
+      "loss": 0.4236,
+      "step": 4460
+    },
+    {
+      "epoch": 0.03575971392228862,
+      "grad_norm": 0.8614292082386968,
+      "learning_rate": 3.57491400687945e-06,
+      "loss": 0.4213,
+      "step": 4470
+    },
+    {
+      "epoch": 0.03583971328229374,
+      "grad_norm": 0.6673883560889204,
+      "learning_rate": 3.582913366930646e-06,
+      "loss": 0.4075,
+      "step": 4480
+    },
+    {
+      "epoch": 0.03591971264229886,
+      "grad_norm": 0.9643089999372365,
+      "learning_rate": 3.5909127269818416e-06,
+      "loss": 0.4546,
+      "step": 4490
+    },
+    {
+      "epoch": 0.03599971200230398,
+      "grad_norm": 0.9297826294459585,
+      "learning_rate": 3.5989120870330375e-06,
+      "loss": 0.4172,
+      "step": 4500
+    },
+    {
+      "epoch": 0.0360797113623091,
+      "grad_norm": 0.7557684985300384,
+      "learning_rate": 3.606911447084234e-06,
+      "loss": 0.4105,
+      "step": 4510
+    },
+    {
+      "epoch": 0.03615971072231422,
+      "grad_norm": 0.7931119899775526,
+      "learning_rate": 3.6149108071354294e-06,
+      "loss": 0.4123,
+      "step": 4520
+    },
+    {
+      "epoch": 0.03623971008231934,
+      "grad_norm": 0.797325353191407,
+      "learning_rate": 3.6229101671866253e-06,
+      "loss": 0.4347,
+      "step": 4530
+    },
+    {
+      "epoch": 0.03631970944232446,
+      "grad_norm": 0.74214327307857,
+      "learning_rate": 3.6309095272378216e-06,
+      "loss": 0.4212,
+      "step": 4540
+    },
+    {
+      "epoch": 0.03639970880232958,
+      "grad_norm": 0.7681829935013766,
+      "learning_rate": 3.638908887289017e-06,
+      "loss": 0.4049,
+      "step": 4550
+    },
+    {
+      "epoch": 0.0364797081623347,
+      "grad_norm": 0.8110044800452652,
+      "learning_rate": 3.646908247340213e-06,
+      "loss": 0.4152,
+      "step": 4560
+    },
+    {
+      "epoch": 0.03655970752233982,
+      "grad_norm": 0.7505043518350412,
+      "learning_rate": 3.654907607391409e-06,
+      "loss": 0.4284,
+      "step": 4570
+    },
+    {
+      "epoch": 0.03663970688234494,
+      "grad_norm": 0.7877830350035486,
+      "learning_rate": 3.662906967442605e-06,
+      "loss": 0.4329,
+      "step": 4580
+    },
+    {
+      "epoch": 0.03671970624235006,
+      "grad_norm": 0.6812701461168307,
+      "learning_rate": 3.6709063274938008e-06,
+      "loss": 0.4173,
+      "step": 4590
+    },
+    {
+      "epoch": 0.03679970560235518,
+      "grad_norm": 0.8001631177482795,
+      "learning_rate": 3.6789056875449967e-06,
+      "loss": 0.4331,
+      "step": 4600
+    },
+    {
+      "epoch": 0.0368797049623603,
+      "grad_norm": 0.7895666869661798,
+      "learning_rate": 3.6869050475961926e-06,
+      "loss": 0.4033,
+      "step": 4610
+    },
+    {
+      "epoch": 0.03695970432236542,
+      "grad_norm": 0.8173983494560131,
+      "learning_rate": 3.6949044076473885e-06,
+      "loss": 0.4164,
+      "step": 4620
+    },
+    {
+      "epoch": 0.03703970368237054,
+      "grad_norm": 0.8378313154048689,
+      "learning_rate": 3.7029037676985844e-06,
+      "loss": 0.3991,
+      "step": 4630
+    },
+    {
+      "epoch": 0.03711970304237566,
+      "grad_norm": 0.9935623637553458,
+      "learning_rate": 3.7109031277497804e-06,
+      "loss": 0.4162,
+      "step": 4640
+    },
+    {
+      "epoch": 0.03719970240238078,
+      "grad_norm": 0.8245648280013409,
+      "learning_rate": 3.7189024878009763e-06,
+      "loss": 0.4269,
+      "step": 4650
+    },
+    {
+      "epoch": 0.0372797017623859,
+      "grad_norm": 0.8268507152029221,
+      "learning_rate": 3.726901847852172e-06,
+      "loss": 0.4169,
+      "step": 4660
+    },
+    {
+      "epoch": 0.03735970112239102,
+      "grad_norm": 0.7781820643589694,
+      "learning_rate": 3.734901207903368e-06,
+      "loss": 0.4277,
+      "step": 4670
+    },
+    {
+      "epoch": 0.03743970048239614,
+      "grad_norm": 0.9906594923724112,
+      "learning_rate": 3.7429005679545636e-06,
+      "loss": 0.4402,
+      "step": 4680
+    },
+    {
+      "epoch": 0.03751969984240126,
+      "grad_norm": 0.7088718853595577,
+      "learning_rate": 3.75089992800576e-06,
+      "loss": 0.4065,
+      "step": 4690
+    },
+    {
+      "epoch": 0.037599699202406384,
+      "grad_norm": 0.7737684662407376,
+      "learning_rate": 3.758899288056956e-06,
+      "loss": 0.43,
+      "step": 4700
+    },
+    {
+      "epoch": 0.037679698562411504,
+      "grad_norm": 0.6799556259390402,
+      "learning_rate": 3.7668986481081513e-06,
+      "loss": 0.4269,
+      "step": 4710
+    },
+    {
+      "epoch": 0.037759697922416624,
+      "grad_norm": 0.8353980941780943,
+      "learning_rate": 3.7748980081593477e-06,
+      "loss": 0.4292,
+      "step": 4720
+    },
+    {
+      "epoch": 0.03783969728242174,
+      "grad_norm": 0.9240909978256971,
+      "learning_rate": 3.7828973682105436e-06,
+      "loss": 0.4218,
+      "step": 4730
+    },
+    {
+      "epoch": 0.03791969664242686,
+      "grad_norm": 0.793600464236656,
+      "learning_rate": 3.790896728261739e-06,
+      "loss": 0.4066,
+      "step": 4740
+    },
+    {
+      "epoch": 0.03799969600243198,
+      "grad_norm": 0.853623661723338,
+      "learning_rate": 3.7988960883129354e-06,
+      "loss": 0.3968,
+      "step": 4750
+    },
+    {
+      "epoch": 0.0380796953624371,
+      "grad_norm": 0.9308117082987841,
+      "learning_rate": 3.8068954483641313e-06,
+      "loss": 0.4127,
+      "step": 4760
+    },
+    {
+      "epoch": 0.03815969472244222,
+      "grad_norm": 0.767583639173416,
+      "learning_rate": 3.814894808415327e-06,
+      "loss": 0.405,
+      "step": 4770
+    },
+    {
+      "epoch": 0.03823969408244734,
+      "grad_norm": 0.7248396423731486,
+      "learning_rate": 3.822894168466522e-06,
+      "loss": 0.4049,
+      "step": 4780
+    },
+    {
+      "epoch": 0.03831969344245246,
+      "grad_norm": 0.7835674270723467,
+      "learning_rate": 3.830893528517719e-06,
+      "loss": 0.4447,
+      "step": 4790
+    },
+    {
+      "epoch": 0.03839969280245758,
+      "grad_norm": 0.7336637502161489,
+      "learning_rate": 3.838892888568915e-06,
+      "loss": 0.415,
+      "step": 4800
+    },
+    {
+      "epoch": 0.0384796921624627,
+      "grad_norm": 0.8211434186162638,
+      "learning_rate": 3.8468922486201105e-06,
+      "loss": 0.4073,
+      "step": 4810
+    },
+    {
+      "epoch": 0.03855969152246782,
+      "grad_norm": 0.6895255239007041,
+      "learning_rate": 3.854891608671307e-06,
+      "loss": 0.4315,
+      "step": 4820
+    },
+    {
+      "epoch": 0.03863969088247294,
+      "grad_norm": 0.8404363403282458,
+      "learning_rate": 3.862890968722502e-06,
+      "loss": 0.4041,
+      "step": 4830
+    },
+    {
+      "epoch": 0.03871969024247806,
+      "grad_norm": 0.8233254492362007,
+      "learning_rate": 3.870890328773698e-06,
+      "loss": 0.4197,
+      "step": 4840
+    },
+    {
+      "epoch": 0.03879968960248318,
+      "grad_norm": 0.7092426325090834,
+      "learning_rate": 3.878889688824894e-06,
+      "loss": 0.415,
+      "step": 4850
+    },
+    {
+      "epoch": 0.0388796889624883,
+      "grad_norm": 0.8335378303937002,
+      "learning_rate": 3.8868890488760905e-06,
+      "loss": 0.4213,
+      "step": 4860
+    },
+    {
+      "epoch": 0.03895968832249342,
+      "grad_norm": 1.0339208338921473,
+      "learning_rate": 3.894888408927286e-06,
+      "loss": 0.4241,
+      "step": 4870
+    },
+    {
+      "epoch": 0.03903968768249854,
+      "grad_norm": 0.7027107330027197,
+      "learning_rate": 3.902887768978482e-06,
+      "loss": 0.4335,
+      "step": 4880
+    },
+    {
+      "epoch": 0.03911968704250366,
+      "grad_norm": 0.6792413425728526,
+      "learning_rate": 3.910887129029678e-06,
+      "loss": 0.4,
+      "step": 4890
+    },
+    {
+      "epoch": 0.03919968640250878,
+      "grad_norm": 0.8489275951527085,
+      "learning_rate": 3.918886489080873e-06,
+      "loss": 0.4207,
+      "step": 4900
+    },
+    {
+      "epoch": 0.0392796857625139,
+      "grad_norm": 0.7357801876777263,
+      "learning_rate": 3.92688584913207e-06,
+      "loss": 0.3994,
+      "step": 4910
+    },
+    {
+      "epoch": 0.03935968512251902,
+      "grad_norm": 0.7141792707926279,
+      "learning_rate": 3.934885209183266e-06,
+      "loss": 0.4154,
+      "step": 4920
+    },
+    {
+      "epoch": 0.03943968448252414,
+      "grad_norm": 0.6915855315198979,
+      "learning_rate": 3.9428845692344615e-06,
+      "loss": 0.433,
+      "step": 4930
+    },
+    {
+      "epoch": 0.03951968384252926,
+      "grad_norm": 0.746029310002541,
+      "learning_rate": 3.950883929285658e-06,
+      "loss": 0.4239,
+      "step": 4940
+    },
+    {
+      "epoch": 0.03959968320253438,
+      "grad_norm": 0.7613311051113969,
+      "learning_rate": 3.958883289336853e-06,
+      "loss": 0.4128,
+      "step": 4950
+    },
+    {
+      "epoch": 0.0396796825625395,
+      "grad_norm": 0.7738245574431311,
+      "learning_rate": 3.966882649388049e-06,
+      "loss": 0.3933,
+      "step": 4960
+    },
+    {
+      "epoch": 0.03975968192254462,
+      "grad_norm": 0.775290020753402,
+      "learning_rate": 3.974882009439245e-06,
+      "loss": 0.4193,
+      "step": 4970
+    },
+    {
+      "epoch": 0.03983968128254974,
+      "grad_norm": 0.6385875554512914,
+      "learning_rate": 3.9828813694904415e-06,
+      "loss": 0.4327,
+      "step": 4980
+    },
+    {
+      "epoch": 0.03991968064255486,
+      "grad_norm": 0.8273897543410583,
+      "learning_rate": 3.990880729541637e-06,
+      "loss": 0.3948,
+      "step": 4990
+    },
+    {
+      "epoch": 0.03999968000255998,
+      "grad_norm": 0.8086324970076214,
+      "learning_rate": 3.998880089592833e-06,
+      "loss": 0.431,
+      "step": 5000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 125001,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 289047165403136.0,
+  "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:b2d0218b2de978c0fb669a44ee4979e8911e040543933e825b0f1ddf72ee6fdf
+size 7736

zero_to_fp32.py

@@ -0,0 +1,760 @@
+#!/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 gc
+import json
+import numpy as np
+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, weights_only=False)
+
+        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 tqdm(files, desc='Loading checkpoint shards'):
+        state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
+        # 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}")
+
+    fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] 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")
+
+
+class GatheredTensor:
+    """
+    A pseudo tensor that collects partitioned weights.
+    It is more memory efficient when there are multiple groups.
+    """
+
+    def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
+        self.flat_groups = flat_groups
+        self.flat_groups_offset = flat_groups_offset
+        self.offset = offset
+        self.partitioned_numel = partitioned_numel
+        self.shape = shape
+        self.dtype = self.flat_groups[0][0].dtype
+
+    def contiguous(self):
+        """
+        Merge partitioned weights from flat_groups into a single tensor.
+        """
+        end_idx = self.offset + self.partitioned_numel
+        world_size = len(self.flat_groups)
+        pad_flat_param_chunks = []
+
+        for rank_i in range(world_size):
+            # for each rank, we need to collect weights from related group/groups
+            flat_groups_at_rank_i = self.flat_groups[rank_i]
+            start_group_id = None
+            end_group_id = None
+            for group_id in range(len(self.flat_groups_offset)):
+                if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
+                    start_group_id = group_id
+                if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
+                    end_group_id = group_id
+                    break
+            # collect weights from related group/groups
+            for group_id in range(start_group_id, end_group_id + 1):
+                flat_tensor = flat_groups_at_rank_i[group_id]
+                start_offset = self.offset - self.flat_groups_offset[group_id]
+                end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
+                pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
+
+        # collect weights from all ranks
+        pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
+        param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
+        return param
+
+
+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 = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * 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
+    flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[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}"
+            )
+
+        # memory efficient tensor
+        tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
+        state_dict[name] = tensor
+        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 to_torch_tensor(state_dict, return_empty_tensor=False):
+    """
+    Convert state_dict of GatheredTensor to torch tensor
+    """
+    torch_state_dict = {}
+    converted_tensors = {}
+    for name, tensor in state_dict.items():
+        tensor_id = id(tensor)
+        if tensor_id in converted_tensors:  # shared tensors
+            shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
+            torch_state_dict[name] = shared_tensor
+        else:
+            converted_tensors[tensor_id] = name
+            if return_empty_tensor:
+                torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
+            else:
+                torch_state_dict[name] = tensor.contiguous()
+    return torch_state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                             tag=None,
+                                             exclude_frozen_parameters=False,
+                                             lazy_mode=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
+        - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
+          Convert the pesduo tensor to torch tensor by ``.contiguous()``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    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.
+
+    Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
+    You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint. Or you can load state_dict in lazy mode ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
+        for name, lazy_tensor in state_dict.item():
+            tensor = lazy_tensor.contiguous()  # to cpu
+            print(name, tensor)
+            # del tensor to release memory if it no longer in use
+    """
+    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")
+
+    state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+    if lazy_mode:
+        return state_dict
+    else:
+        return to_torch_tensor(state_dict)
+
+
+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,
+                                                          lazy_mode=True)
+
+    # 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")
+        # an memory-efficient approach for sharding
+        empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
+        state_dict_split = split_torch_state_dict_into_shards(empty_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 by shard
+    os.makedirs(output_dir, exist_ok=True)
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
+        shard_state_dict = to_torch_tensor(shard_state_dict)
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard_state_dict, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard_state_dict, output_path)
+        # release the memory of current shard
+        for tensor_name in list(shard_state_dict.keys()):
+            del state_dict[tensor_name]
+            del shard_state_dict[tensor_name]
+        del shard_state_dict
+        gc.collect()
+
+    # 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)