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

.ipynb_checkpoints/README-checkpoint.md

@@ -0,0 +1,69 @@
+---
+library_name: transformers
+license: other
+base_model: Qwen3.5-9B
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: qwen35_caption_galore
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# qwen35_caption_galore
+
+This model is a fine-tuned version of [/workspace/models/Qwen3.5-9B](https://huggingface.co//workspace/models/Qwen3.5-9B) on the my_caption dataset.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- family_to_muon_lr = {
+        "language": _fallback(getattr(training_args, "language_muon_lr", 2e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_muon_lr", 5e-5), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_muon_lr", 5e-5), merger_lr),
+    }
+
+    family_to_adamw_lr = {
+        "language": _fallback(getattr(training_args, "language_adamw_lr", 1e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_adamw_lr", 5e-6), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_adamw_lr", 1e-5), merger_lr),
+    }
+- train_batch_size: 3
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- gradient_accumulation_steps: 20
+- total_train_batch_size: 60
+- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine_with_min_lr
+- lr_scheduler_warmup_steps: 0.05
+- num_epochs: 4
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 5.5.3
+- Pytorch 2.11.0+cu128
+- Datasets 4.0.0
+- Tokenizers 0.22.2

README.md

@@ -0,0 +1,69 @@
+---
+library_name: transformers
+license: other
+base_model: Qwen3.5-9B
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: qwen35_caption_galore
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# qwen35_caption_galore
+
+This model is a fine-tuned version of [/workspace/models/Qwen3.5-9B](https://huggingface.co//workspace/models/Qwen3.5-9B) on the my_caption dataset.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- family_to_muon_lr = {
+        "language": _fallback(getattr(training_args, "language_muon_lr", 2e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_muon_lr", 5e-5), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_muon_lr", 5e-5), merger_lr),
+    }
+
+    family_to_adamw_lr = {
+        "language": _fallback(getattr(training_args, "language_adamw_lr", 1e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_adamw_lr", 5e-6), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_adamw_lr", 1e-5), merger_lr),
+    }
+- train_batch_size: 3
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- gradient_accumulation_steps: 20
+- total_train_batch_size: 60
+- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine_with_min_lr
+- lr_scheduler_warmup_steps: 0.05
+- num_epochs: 4
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 5.5.3
+- Pytorch 2.11.0+cu128
+- Datasets 4.0.0
+- Tokenizers 0.22.2

all_results.json

@@ -0,0 +1,9 @@
+{
+    "effective_tokens_per_sec": 7652.111502010357,
+    "epoch": 4.0,
+    "total_flos": 3.519329211045118e+18,
+    "train_loss": 0.8036329971025308,
+    "train_runtime": 8899.7642,
+    "train_samples_per_second": 6.769,
+    "train_steps_per_second": 0.113
+}

chat_template.jinja

@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+    {%- if content is string %}
+        {{- content }}
+    {%- elif content is iterable and content is not mapping %}
+        {%- for item in content %}
+            {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+                {%- if is_system_content %}
+                    {{- raise_exception('System message cannot contain images.') }}
+                {%- endif %}
+                {%- if do_vision_count %}
+                    {%- set image_count.value = image_count.value + 1 %}
+                {%- endif %}
+                {%- if add_vision_id %}
+                    {{- 'Picture ' ~ image_count.value ~ ': ' }}
+                {%- endif %}
+                {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+            {%- elif 'video' in item or item.type == 'video' %}
+                {%- if is_system_content %}
+                    {{- raise_exception('System message cannot contain videos.') }}
+                {%- endif %}
+                {%- if do_vision_count %}
+                    {%- set video_count.value = video_count.value + 1 %}
+                {%- endif %}
+                {%- if add_vision_id %}
+                    {{- 'Video ' ~ video_count.value ~ ': ' }}
+                {%- endif %}
+                {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+            {%- elif 'text' in item %}
+                {{- item.text }}
+            {%- else %}
+                {{- raise_exception('Unexpected item type in content.') }}
+            {%- endif %}
+        {%- endfor %}
+    {%- elif content is none or content is undefined %}
+        {{- '' }}
+    {%- else %}
+        {{- raise_exception('Unexpected content type.') }}
+    {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+    {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+    {{- '<|im_start|>system\n' }}
+    {{- "# Tools\n\nYou have access to the following functions:\n\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>" }}
+    {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n<tool_call>\n<function=example_function_name>\n<parameter=example_parameter_1>\nvalue_1\n</parameter>\n<parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n</parameter>\n</function>\n</tool_call>\n\n<IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner <function=...></function> block must be nested within <tool_call></tool_call> XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n</IMPORTANT>' }}
+    {%- if messages[0].role == 'system' %}
+        {%- set content = render_content(messages[0].content, false, true)|trim %}
+        {%- if content %}
+            {{- '\n\n' + content }}
+        {%- endif %}
+    {%- endif %}
+    {{- '<|im_end|>\n' }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {%- set content = render_content(messages[0].content, false, true)|trim %}
+        {{- '<|im_start|>system\n' + 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" %}
+        {%- set content = render_content(message.content, false)|trim %}
+        {%- if not(content.startswith('<tool_response>') and content.endswith('</tool_response>')) %}
+            {%- set ns.multi_step_tool = false %}
+            {%- set ns.last_query_index = index %}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+    {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+    {%- set content = render_content(message.content, true)|trim %}
+    {%- if message.role == "system" %}
+        {%- if not loop.first %}
+            {{- raise_exception('System message must be at the beginning.') }}
+        {%- endif %}
+    {%- elif message.role == "user" %}
+        {{- '<|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 %}
+        {%- set reasoning_content = reasoning_content|trim %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content + '\n</think>\n\n' + content }}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if tool_call.function is defined %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {%- if loop.first %}
+                    {%- if content|trim %}
+                        {{- '\n\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                    {%- else %}
+                        {{- '<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                    {%- endif %}
+                {%- else %}
+                    {{- '\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                {%- endif %}
+                {%- if tool_call.arguments is defined %}
+                    {%- for args_name, args_value in tool_call.arguments|items %}
+                        {{- '<parameter=' + args_name + '>\n' }}
+                        {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %}
+                        {{- args_value }}
+                        {{- '\n</parameter>\n' }}
+                    {%- endfor %}
+                {%- endif %}
+                {{- '</function>\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.previtem and loop.previtem.role != "tool" %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- content }}
+        {{- '\n</tool_response>' }}
+        {%- if not loop.last and loop.nextitem.role != "tool" %}
+            {{- '<|im_end|>\n' }}
+        {%- elif loop.last %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- else %}
+        {{- raise_exception('Unexpected message role.') }}
+    {%- 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' }}
+    {%- else %}
+        {{- '<think>\n' }}
+    {%- endif %}
+{%- endif %}

config.json

@@ -0,0 +1,113 @@
+{
+  "architectures": [
+    "Qwen3_5ForConditionalGeneration"
+  ],
+  "dtype": "bfloat16",
+  "eos_token_id": 248046,
+  "hidden_size": 4096,
+  "image_token_id": 248056,
+  "model_type": "qwen3_5",
+  "pad_token_id": 248044,
+  "text_config": {
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "attn_output_gate": true,
+    "bos_token_id": null,
+    "dtype": "bfloat16",
+    "eos_token_id": 248044,
+    "full_attention_interval": 4,
+    "head_dim": 256,
+    "hidden_act": "silu",
+    "hidden_size": 4096,
+    "initializer_range": 0.02,
+    "intermediate_size": 12288,
+    "layer_types": [
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention"
+    ],
+    "linear_conv_kernel_dim": 4,
+    "linear_key_head_dim": 128,
+    "linear_num_key_heads": 16,
+    "linear_num_value_heads": 32,
+    "linear_value_head_dim": 128,
+    "mamba_ssm_dtype": "float32",
+    "max_position_embeddings": 262144,
+    "mlp_only_layers": [],
+    "model_type": "qwen3_5_text",
+    "mtp_num_hidden_layers": 1,
+    "mtp_use_dedicated_embeddings": false,
+    "num_attention_heads": 16,
+    "num_hidden_layers": 32,
+    "num_key_value_heads": 4,
+    "pad_token_id": null,
+    "partial_rotary_factor": 0.25,
+    "rms_norm_eps": 1e-06,
+    "rope_parameters": {
+      "mrope_interleaved": true,
+      "mrope_section": [
+        11,
+        11,
+        10
+      ],
+      "partial_rotary_factor": 0.25,
+      "rope_theta": 10000000,
+      "rope_type": "default"
+    },
+    "tie_word_embeddings": false,
+    "use_cache": false,
+    "vocab_size": 248320
+  },
+  "tie_word_embeddings": false,
+  "transformers_version": "5.5.3",
+  "use_cache": false,
+  "video_token_id": 248057,
+  "vision_config": {
+    "deepstack_visual_indexes": [],
+    "depth": 27,
+    "dtype": "bfloat16",
+    "hidden_act": "gelu_pytorch_tanh",
+    "hidden_size": 1152,
+    "in_channels": 3,
+    "initializer_range": 0.02,
+    "intermediate_size": 4304,
+    "model_type": "qwen3_5",
+    "num_heads": 16,
+    "num_position_embeddings": 2304,
+    "out_hidden_size": 4096,
+    "patch_size": 16,
+    "spatial_merge_size": 2,
+    "temporal_patch_size": 2
+  },
+  "vision_end_token_id": 248054,
+  "vision_start_token_id": 248053
+}

generation_config.json

@@ -0,0 +1,10 @@
+{
+  "_from_model_config": true,
+  "eos_token_id": [
+    248046,
+    248044
+  ],
+  "pad_token_id": 248044,
+  "transformers_version": "5.5.3",
+  "use_cache": true
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3f77dcf06f82d128139550e7b30ff2583178f7d1a6610b8756a4bcdf12c39bb
+size 20859273368

processor_config.json

@@ -0,0 +1,60 @@
+{
+  "image_processor": {
+    "do_convert_rgb": true,
+    "do_normalize": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "image_mean": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "image_processor_type": "Qwen2VLImageProcessor",
+    "image_std": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "merge_size": 2,
+    "patch_size": 16,
+    "resample": 3,
+    "rescale_factor": 0.00392156862745098,
+    "size": {
+      "longest_edge": 16777216,
+      "shortest_edge": 65536
+    },
+    "temporal_patch_size": 2
+  },
+  "processor_class": "Qwen3VLProcessor",
+  "video_processor": {
+    "do_convert_rgb": true,
+    "do_normalize": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "do_sample_frames": true,
+    "fps": 2,
+    "image_mean": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "image_std": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "max_frames": 768,
+    "merge_size": 2,
+    "min_frames": 4,
+    "patch_size": 16,
+    "resample": 3,
+    "rescale_factor": 0.00392156862745098,
+    "return_metadata": false,
+    "size": {
+      "longest_edge": 25165824,
+      "shortest_edge": 4096
+    },
+    "temporal_patch_size": 2,
+    "video_processor_type": "Qwen3VLVideoProcessor"
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87a7830d63fcf43bf241c3c5242e96e62dd3fdc29224ca26fed8ea333db72de4
+size 19989343

tokenizer_config.json

@@ -0,0 +1,33 @@
+{
+  "add_prefix_space": false,
+  "audio_bos_token": "<|audio_start|>",
+  "audio_eos_token": "<|audio_end|>",
+  "audio_token": "<|audio_pad|>",
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "image_token": "<|image_pad|>",
+  "is_local": true,
+  "model_max_length": 262144,
+  "model_specific_special_tokens": {
+    "audio_bos_token": "<|audio_start|>",
+    "audio_eos_token": "<|audio_end|>",
+    "audio_token": "<|audio_pad|>",
+    "image_token": "<|image_pad|>",
+    "video_token": "<|video_pad|>",
+    "vision_bos_token": "<|vision_start|>",
+    "vision_eos_token": "<|vision_end|>"
+  },
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+  "processor_class": "Qwen3VLProcessor",
+  "split_special_tokens": false,
+  "tokenizer_class": "TokenizersBackend",
+  "unk_token": null,
+  "video_token": "<|video_pad|>",
+  "vision_bos_token": "<|vision_start|>",
+  "vision_eos_token": "<|vision_end|>"
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "effective_tokens_per_sec": 7652.111502010357,
+    "epoch": 4.0,
+    "total_flos": 3.519329211045118e+18,
+    "train_loss": 0.8036329971025308,
+    "train_runtime": 8899.7642,
+    "train_samples_per_second": 6.769,
+    "train_steps_per_second": 0.113
+}

trainer_state.json

@@ -0,0 +1,3557 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 500,
+  "global_step": 1004,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00796812749003984,
+      "grad_norm": 1.1519681215286255,
+      "learning_rate": 1.9607843137254904e-07,
+      "loss": 2.0970637798309326,
+      "step": 2
+    },
+    {
+      "epoch": 0.01593625498007968,
+      "grad_norm": 0.6537962555885315,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 2.013030529022217,
+      "step": 4
+    },
+    {
+      "epoch": 0.02390438247011952,
+      "grad_norm": 2.2287347316741943,
+      "learning_rate": 9.80392156862745e-07,
+      "loss": 2.177837610244751,
+      "step": 6
+    },
+    {
+      "epoch": 0.03187250996015936,
+      "grad_norm": 0.7109684348106384,
+      "learning_rate": 1.3725490196078434e-06,
+      "loss": 1.8894720077514648,
+      "step": 8
+    },
+    {
+      "epoch": 0.0398406374501992,
+      "grad_norm": 5.645593643188477,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 1.4646748304367065,
+      "step": 10
+    },
+    {
+      "epoch": 0.04780876494023904,
+      "grad_norm": 0.7475039958953857,
+      "learning_rate": 2.1568627450980393e-06,
+      "loss": 1.7005442380905151,
+      "step": 12
+    },
+    {
+      "epoch": 0.055776892430278883,
+      "grad_norm": 0.8137855529785156,
+      "learning_rate": 2.549019607843137e-06,
+      "loss": 2.2653238773345947,
+      "step": 14
+    },
+    {
+      "epoch": 0.06374501992031872,
+      "grad_norm": 2.9936726093292236,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 2.1347434520721436,
+      "step": 16
+    },
+    {
+      "epoch": 0.07171314741035857,
+      "grad_norm": 3.7850115299224854,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.6133207082748413,
+      "step": 18
+    },
+    {
+      "epoch": 0.0796812749003984,
+      "grad_norm": 0.582400918006897,
+      "learning_rate": 3.7254901960784316e-06,
+      "loss": 1.5486186742782593,
+      "step": 20
+    },
+    {
+      "epoch": 0.08764940239043825,
+      "grad_norm": 0.4687569737434387,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 1.1839547157287598,
+      "step": 22
+    },
+    {
+      "epoch": 0.09561752988047809,
+      "grad_norm": 0.5278186202049255,
+      "learning_rate": 4.509803921568628e-06,
+      "loss": 1.3662710189819336,
+      "step": 24
+    },
+    {
+      "epoch": 0.10358565737051793,
+      "grad_norm": 0.3190319836139679,
+      "learning_rate": 4.901960784313726e-06,
+      "loss": 1.4518065452575684,
+      "step": 26
+    },
+    {
+      "epoch": 0.11155378486055777,
+      "grad_norm": 0.4343985319137573,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 1.6269359588623047,
+      "step": 28
+    },
+    {
+      "epoch": 0.11952191235059761,
+      "grad_norm": 0.6612064838409424,
+      "learning_rate": 5.686274509803922e-06,
+      "loss": 1.3382073640823364,
+      "step": 30
+    },
+    {
+      "epoch": 0.12749003984063745,
+      "grad_norm": 0.5706899762153625,
+      "learning_rate": 6.07843137254902e-06,
+      "loss": 1.1689900159835815,
+      "step": 32
+    },
+    {
+      "epoch": 0.13545816733067728,
+      "grad_norm": 0.5469005107879639,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 1.2483381032943726,
+      "step": 34
+    },
+    {
+      "epoch": 0.14342629482071714,
+      "grad_norm": 0.36703023314476013,
+      "learning_rate": 6.862745098039216e-06,
+      "loss": 1.4096142053604126,
+      "step": 36
+    },
+    {
+      "epoch": 0.15139442231075698,
+      "grad_norm": 1.020111322402954,
+      "learning_rate": 7.2549019607843145e-06,
+      "loss": 1.195603847503662,
+      "step": 38
+    },
+    {
+      "epoch": 0.1593625498007968,
+      "grad_norm": 0.31180691719055176,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 1.3204171657562256,
+      "step": 40
+    },
+    {
+      "epoch": 0.16733067729083664,
+      "grad_norm": 0.9232037663459778,
+      "learning_rate": 8.03921568627451e-06,
+      "loss": 0.7868313193321228,
+      "step": 42
+    },
+    {
+      "epoch": 0.1752988047808765,
+      "grad_norm": 0.6603556871414185,
+      "learning_rate": 8.43137254901961e-06,
+      "loss": 0.9809900522232056,
+      "step": 44
+    },
+    {
+      "epoch": 0.18326693227091634,
+      "grad_norm": 1.6366173028945923,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.9468201994895935,
+      "step": 46
+    },
+    {
+      "epoch": 0.19123505976095617,
+      "grad_norm": 0.5651039481163025,
+      "learning_rate": 9.215686274509804e-06,
+      "loss": 1.1833889484405518,
+      "step": 48
+    },
+    {
+      "epoch": 0.199203187250996,
+      "grad_norm": 0.35958272218704224,
+      "learning_rate": 9.607843137254903e-06,
+      "loss": 1.3381084203720093,
+      "step": 50
+    },
+    {
+      "epoch": 0.20717131474103587,
+      "grad_norm": 0.6453939080238342,
+      "learning_rate": 1e-05,
+      "loss": 1.2928056716918945,
+      "step": 52
+    },
+    {
+      "epoch": 0.2151394422310757,
+      "grad_norm": 2.1719014644622803,
+      "learning_rate": 9.999902196391087e-06,
+      "loss": 1.09474778175354,
+      "step": 54
+    },
+    {
+      "epoch": 0.22310756972111553,
+      "grad_norm": 0.5366278886795044,
+      "learning_rate": 9.999608789815699e-06,
+      "loss": 1.5230549573898315,
+      "step": 56
+    },
+    {
+      "epoch": 0.23107569721115537,
+      "grad_norm": 0.42068353295326233,
+      "learning_rate": 9.999119793027713e-06,
+      "loss": 1.3382229804992676,
+      "step": 58
+    },
+    {
+      "epoch": 0.23904382470119523,
+      "grad_norm": 0.6967461109161377,
+      "learning_rate": 9.998435227282975e-06,
+      "loss": 0.7599650621414185,
+      "step": 60
+    },
+    {
+      "epoch": 0.24701195219123506,
+      "grad_norm": 0.4790613055229187,
+      "learning_rate": 9.997555122338372e-06,
+      "loss": 1.341271162033081,
+      "step": 62
+    },
+    {
+      "epoch": 0.2549800796812749,
+      "grad_norm": 0.6073594689369202,
+      "learning_rate": 9.996479516450544e-06,
+      "loss": 1.1212892532348633,
+      "step": 64
+    },
+    {
+      "epoch": 0.26294820717131473,
+      "grad_norm": 0.48113441467285156,
+      "learning_rate": 9.99520845637422e-06,
+      "loss": 1.2697763442993164,
+      "step": 66
+    },
+    {
+      "epoch": 0.27091633466135456,
+      "grad_norm": 0.5884555578231812,
+      "learning_rate": 9.993741997360185e-06,
+      "loss": 0.8300902247428894,
+      "step": 68
+    },
+    {
+      "epoch": 0.2788844621513944,
+      "grad_norm": 0.7450176477432251,
+      "learning_rate": 9.992080203152872e-06,
+      "loss": 1.5441479682922363,
+      "step": 70
+    },
+    {
+      "epoch": 0.2868525896414343,
+      "grad_norm": 1.1038637161254883,
+      "learning_rate": 9.990223145987608e-06,
+      "loss": 0.8233323097229004,
+      "step": 72
+    },
+    {
+      "epoch": 0.2948207171314741,
+      "grad_norm": 8.860465049743652,
+      "learning_rate": 9.98817090658745e-06,
+      "loss": 1.1380279064178467,
+      "step": 74
+    },
+    {
+      "epoch": 0.30278884462151395,
+      "grad_norm": 1.0202000141143799,
+      "learning_rate": 9.985923574159696e-06,
+      "loss": 1.4511159658432007,
+      "step": 76
+    },
+    {
+      "epoch": 0.3107569721115538,
+      "grad_norm": 2.2417469024658203,
+      "learning_rate": 9.983481246392003e-06,
+      "loss": 0.912439227104187,
+      "step": 78
+    },
+    {
+      "epoch": 0.3187250996015936,
+      "grad_norm": 0.3263985216617584,
+      "learning_rate": 9.980844029448135e-06,
+      "loss": 0.8763936758041382,
+      "step": 80
+    },
+    {
+      "epoch": 0.32669322709163345,
+      "grad_norm": 0.42795252799987793,
+      "learning_rate": 9.97801203796335e-06,
+      "loss": 1.3310266733169556,
+      "step": 82
+    },
+    {
+      "epoch": 0.3346613545816733,
+      "grad_norm": 1.0190571546554565,
+      "learning_rate": 9.974985395039423e-06,
+      "loss": 0.9474778175354004,
+      "step": 84
+    },
+    {
+      "epoch": 0.3426294820717131,
+      "grad_norm": 1.1264762878417969,
+      "learning_rate": 9.971764232239287e-06,
+      "loss": 1.3073441982269287,
+      "step": 86
+    },
+    {
+      "epoch": 0.350597609561753,
+      "grad_norm": 1.365116000175476,
+      "learning_rate": 9.968348689581315e-06,
+      "loss": 0.9702357053756714,
+      "step": 88
+    },
+    {
+      "epoch": 0.35856573705179284,
+      "grad_norm": 1.018405556678772,
+      "learning_rate": 9.964738915533242e-06,
+      "loss": 1.317631721496582,
+      "step": 90
+    },
+    {
+      "epoch": 0.3665338645418327,
+      "grad_norm": 1.190874695777893,
+      "learning_rate": 9.960935067005706e-06,
+      "loss": 1.1108540296554565,
+      "step": 92
+    },
+    {
+      "epoch": 0.3745019920318725,
+      "grad_norm": 0.6522363424301147,
+      "learning_rate": 9.956937309345418e-06,
+      "loss": 1.2383490800857544,
+      "step": 94
+    },
+    {
+      "epoch": 0.38247011952191234,
+      "grad_norm": 0.7052263021469116,
+      "learning_rate": 9.952745816327998e-06,
+      "loss": 1.0969046354293823,
+      "step": 96
+    },
+    {
+      "epoch": 0.3904382470119522,
+      "grad_norm": 0.48337286710739136,
+      "learning_rate": 9.94836077015039e-06,
+      "loss": 1.2495427131652832,
+      "step": 98
+    },
+    {
+      "epoch": 0.398406374501992,
+      "grad_norm": 1.1672080755233765,
+      "learning_rate": 9.943782361422975e-06,
+      "loss": 0.8611827492713928,
+      "step": 100
+    },
+    {
+      "epoch": 0.4063745019920319,
+      "grad_norm": 0.7386517524719238,
+      "learning_rate": 9.939010789161259e-06,
+      "loss": 1.264111876487732,
+      "step": 102
+    },
+    {
+      "epoch": 0.41434262948207173,
+      "grad_norm": 0.25491490960121155,
+      "learning_rate": 9.934046260777239e-06,
+      "loss": 0.798167884349823,
+      "step": 104
+    },
+    {
+      "epoch": 0.42231075697211157,
+      "grad_norm": 1.0984123945236206,
+      "learning_rate": 9.928888992070373e-06,
+      "loss": 0.7865558862686157,
+      "step": 106
+    },
+    {
+      "epoch": 0.4302788844621514,
+      "grad_norm": 0.7432213425636292,
+      "learning_rate": 9.923539207218219e-06,
+      "loss": 0.9980818629264832,
+      "step": 108
+    },
+    {
+      "epoch": 0.43824701195219123,
+      "grad_norm": 0.34016844630241394,
+      "learning_rate": 9.91799713876667e-06,
+      "loss": 1.2785511016845703,
+      "step": 110
+    },
+    {
+      "epoch": 0.44621513944223107,
+      "grad_norm": 1.4738695621490479,
+      "learning_rate": 9.912263027619856e-06,
+      "loss": 0.8448340892791748,
+      "step": 112
+    },
+    {
+      "epoch": 0.4541832669322709,
+      "grad_norm": 1.327816367149353,
+      "learning_rate": 9.906337123029675e-06,
+      "loss": 1.134405255317688,
+      "step": 114
+    },
+    {
+      "epoch": 0.46215139442231074,
+      "grad_norm": 0.4198015332221985,
+      "learning_rate": 9.90021968258495e-06,
+      "loss": 1.2850412130355835,
+      "step": 116
+    },
+    {
+      "epoch": 0.4701195219123506,
+      "grad_norm": 0.29487475752830505,
+      "learning_rate": 9.893910972200238e-06,
+      "loss": 1.348087191581726,
+      "step": 118
+    },
+    {
+      "epoch": 0.47808764940239046,
+      "grad_norm": 0.6308404803276062,
+      "learning_rate": 9.887411266104268e-06,
+      "loss": 1.2281802892684937,
+      "step": 120
+    },
+    {
+      "epoch": 0.4860557768924303,
+      "grad_norm": 0.4446815550327301,
+      "learning_rate": 9.880720846828026e-06,
+      "loss": 0.976684033870697,
+      "step": 122
+    },
+    {
+      "epoch": 0.4940239043824701,
+      "grad_norm": 0.6750710010528564,
+      "learning_rate": 9.873840005192466e-06,
+      "loss": 1.0549778938293457,
+      "step": 124
+    },
+    {
+      "epoch": 0.50199203187251,
+      "grad_norm": 4.390827178955078,
+      "learning_rate": 9.866769040295875e-06,
+      "loss": 0.95052170753479,
+      "step": 126
+    },
+    {
+      "epoch": 0.5099601593625498,
+      "grad_norm": 0.35833993554115295,
+      "learning_rate": 9.859508259500869e-06,
+      "loss": 1.2717690467834473,
+      "step": 128
+    },
+    {
+      "epoch": 0.5179282868525896,
+      "grad_norm": 0.34393975138664246,
+      "learning_rate": 9.852057978421033e-06,
+      "loss": 0.9470924139022827,
+      "step": 130
+    },
+    {
+      "epoch": 0.5258964143426295,
+      "grad_norm": 0.5196332931518555,
+      "learning_rate": 9.844418520907198e-06,
+      "loss": 0.7987101078033447,
+      "step": 132
+    },
+    {
+      "epoch": 0.5338645418326693,
+      "grad_norm": 0.3912753164768219,
+      "learning_rate": 9.836590219033375e-06,
+      "loss": 0.9098261594772339,
+      "step": 134
+    },
+    {
+      "epoch": 0.5418326693227091,
+      "grad_norm": 0.548719584941864,
+      "learning_rate": 9.828573413082306e-06,
+      "loss": 1.2886285781860352,
+      "step": 136
+    },
+    {
+      "epoch": 0.549800796812749,
+      "grad_norm": 0.8654932379722595,
+      "learning_rate": 9.820368451530681e-06,
+      "loss": 0.940203070640564,
+      "step": 138
+    },
+    {
+      "epoch": 0.5577689243027888,
+      "grad_norm": 2.5796964168548584,
+      "learning_rate": 9.811975691033992e-06,
+      "loss": 1.0959604978561401,
+      "step": 140
+    },
+    {
+      "epoch": 0.5657370517928287,
+      "grad_norm": 0.477183997631073,
+      "learning_rate": 9.80339549641102e-06,
+      "loss": 1.0312070846557617,
+      "step": 142
+    },
+    {
+      "epoch": 0.5737051792828686,
+      "grad_norm": 0.3567017912864685,
+      "learning_rate": 9.794628240627992e-06,
+      "loss": 1.2184264659881592,
+      "step": 144
+    },
+    {
+      "epoch": 0.5816733067729084,
+      "grad_norm": 0.4620193839073181,
+      "learning_rate": 9.785674304782349e-06,
+      "loss": 1.2170346975326538,
+      "step": 146
+    },
+    {
+      "epoch": 0.5896414342629482,
+      "grad_norm": 0.395216703414917,
+      "learning_rate": 9.776534078086202e-06,
+      "loss": 0.9650822877883911,
+      "step": 148
+    },
+    {
+      "epoch": 0.5976095617529881,
+      "grad_norm": 0.8579673767089844,
+      "learning_rate": 9.767207957849396e-06,
+      "loss": 0.9703203439712524,
+      "step": 150
+    },
+    {
+      "epoch": 0.6055776892430279,
+      "grad_norm": 0.3818407356739044,
+      "learning_rate": 9.757696349462252e-06,
+      "loss": 0.9532766938209534,
+      "step": 152
+    },
+    {
+      "epoch": 0.6135458167330677,
+      "grad_norm": 0.5531460642814636,
+      "learning_rate": 9.747999666377934e-06,
+      "loss": 1.1107113361358643,
+      "step": 154
+    },
+    {
+      "epoch": 0.6215139442231076,
+      "grad_norm": 2.03843092918396,
+      "learning_rate": 9.738118330094491e-06,
+      "loss": 1.0462383031845093,
+      "step": 156
+    },
+    {
+      "epoch": 0.6294820717131474,
+      "grad_norm": 0.500770092010498,
+      "learning_rate": 9.728052770136518e-06,
+      "loss": 1.1488879919052124,
+      "step": 158
+    },
+    {
+      "epoch": 0.6374501992031872,
+      "grad_norm": 0.279585063457489,
+      "learning_rate": 9.717803424036501e-06,
+      "loss": 1.0134367942810059,
+      "step": 160
+    },
+    {
+      "epoch": 0.6454183266932271,
+      "grad_norm": 0.2347595989704132,
+      "learning_rate": 9.707370737315793e-06,
+      "loss": 0.7832069993019104,
+      "step": 162
+    },
+    {
+      "epoch": 0.6533864541832669,
+      "grad_norm": 0.21960224211215973,
+      "learning_rate": 9.69675516346524e-06,
+      "loss": 1.1398905515670776,
+      "step": 164
+    },
+    {
+      "epoch": 0.6613545816733067,
+      "grad_norm": 1.2041698694229126,
+      "learning_rate": 9.685957163925479e-06,
+      "loss": 0.8306769728660583,
+      "step": 166
+    },
+    {
+      "epoch": 0.6693227091633466,
+      "grad_norm": 0.33941978216171265,
+      "learning_rate": 9.67497720806688e-06,
+      "loss": 1.2318183183670044,
+      "step": 168
+    },
+    {
+      "epoch": 0.6772908366533864,
+      "grad_norm": 2.14397931098938,
+      "learning_rate": 9.663815773169131e-06,
+      "loss": 0.9232658743858337,
+      "step": 170
+    },
+    {
+      "epoch": 0.6852589641434262,
+      "grad_norm": 0.47329750657081604,
+      "learning_rate": 9.652473344400507e-06,
+      "loss": 0.8883454203605652,
+      "step": 172
+    },
+    {
+      "epoch": 0.6932270916334662,
+      "grad_norm": 0.2859004735946655,
+      "learning_rate": 9.640950414796772e-06,
+      "loss": 1.224610447883606,
+      "step": 174
+    },
+    {
+      "epoch": 0.701195219123506,
+      "grad_norm": 0.8783998489379883,
+      "learning_rate": 9.629247485239747e-06,
+      "loss": 0.7565767765045166,
+      "step": 176
+    },
+    {
+      "epoch": 0.7091633466135459,
+      "grad_norm": 1.0212678909301758,
+      "learning_rate": 9.617365064435544e-06,
+      "loss": 0.9688025116920471,
+      "step": 178
+    },
+    {
+      "epoch": 0.7171314741035857,
+      "grad_norm": 1.9316108226776123,
+      "learning_rate": 9.605303668892443e-06,
+      "loss": 1.0620930194854736,
+      "step": 180
+    },
+    {
+      "epoch": 0.7250996015936255,
+      "grad_norm": 0.13066768646240234,
+      "learning_rate": 9.593063822898451e-06,
+      "loss": 0.505932092666626,
+      "step": 182
+    },
+    {
+      "epoch": 0.7330677290836654,
+      "grad_norm": 0.4132096767425537,
+      "learning_rate": 9.580646058498505e-06,
+      "loss": 1.3479992151260376,
+      "step": 184
+    },
+    {
+      "epoch": 0.7410358565737052,
+      "grad_norm": 0.3841014802455902,
+      "learning_rate": 9.56805091547135e-06,
+      "loss": 1.2776392698287964,
+      "step": 186
+    },
+    {
+      "epoch": 0.749003984063745,
+      "grad_norm": 0.43801066279411316,
+      "learning_rate": 9.55527894130607e-06,
+      "loss": 1.11171555519104,
+      "step": 188
+    },
+    {
+      "epoch": 0.7569721115537849,
+      "grad_norm": 1.432983160018921,
+      "learning_rate": 9.542330691178296e-06,
+      "loss": 0.8871626257896423,
+      "step": 190
+    },
+    {
+      "epoch": 0.7649402390438247,
+      "grad_norm": 0.5004543662071228,
+      "learning_rate": 9.529206727926071e-06,
+      "loss": 1.215728759765625,
+      "step": 192
+    },
+    {
+      "epoch": 0.7729083665338645,
+      "grad_norm": 0.33537885546684265,
+      "learning_rate": 9.515907622025376e-06,
+      "loss": 0.9015203714370728,
+      "step": 194
+    },
+    {
+      "epoch": 0.7808764940239044,
+      "grad_norm": 3.0764424800872803,
+      "learning_rate": 9.502433951565347e-06,
+      "loss": 0.5928923487663269,
+      "step": 196
+    },
+    {
+      "epoch": 0.7888446215139442,
+      "grad_norm": 1.1754522323608398,
+      "learning_rate": 9.48878630222314e-06,
+      "loss": 0.7800745964050293,
+      "step": 198
+    },
+    {
+      "epoch": 0.796812749003984,
+      "grad_norm": 1.280935287475586,
+      "learning_rate": 9.474965267238469e-06,
+      "loss": 1.0495020151138306,
+      "step": 200
+    },
+    {
+      "epoch": 0.8047808764940239,
+      "grad_norm": 0.20368656516075134,
+      "learning_rate": 9.460971447387824e-06,
+      "loss": 0.9815849661827087,
+      "step": 202
+    },
+    {
+      "epoch": 0.8127490039840638,
+      "grad_norm": 0.4857024550437927,
+      "learning_rate": 9.446805450958353e-06,
+      "loss": 1.528865933418274,
+      "step": 204
+    },
+    {
+      "epoch": 0.8207171314741036,
+      "grad_norm": 0.30279162526130676,
+      "learning_rate": 9.43246789372142e-06,
+      "loss": 1.2313956022262573,
+      "step": 206
+    },
+    {
+      "epoch": 0.8286852589641435,
+      "grad_norm": 0.3518405556678772,
+      "learning_rate": 9.417959398905846e-06,
+      "loss": 0.6028034687042236,
+      "step": 208
+    },
+    {
+      "epoch": 0.8366533864541833,
+      "grad_norm": 0.4644409120082855,
+      "learning_rate": 9.40328059717081e-06,
+      "loss": 1.274383783340454,
+      "step": 210
+    },
+    {
+      "epoch": 0.8446215139442231,
+      "grad_norm": 0.3207145035266876,
+      "learning_rate": 9.388432126578438e-06,
+      "loss": 1.2243791818618774,
+      "step": 212
+    },
+    {
+      "epoch": 0.852589641434263,
+      "grad_norm": 0.3495004177093506,
+      "learning_rate": 9.373414632566066e-06,
+      "loss": 0.9430820345878601,
+      "step": 214
+    },
+    {
+      "epoch": 0.8605577689243028,
+      "grad_norm": 0.7576279640197754,
+      "learning_rate": 9.35822876791819e-06,
+      "loss": 1.426702618598938,
+      "step": 216
+    },
+    {
+      "epoch": 0.8685258964143426,
+      "grad_norm": 0.7927657961845398,
+      "learning_rate": 9.342875192738083e-06,
+      "loss": 0.8673502802848816,
+      "step": 218
+    },
+    {
+      "epoch": 0.8764940239043825,
+      "grad_norm": 0.6116431951522827,
+      "learning_rate": 9.327354574419109e-06,
+      "loss": 1.2359604835510254,
+      "step": 220
+    },
+    {
+      "epoch": 0.8844621513944223,
+      "grad_norm": 0.4329512119293213,
+      "learning_rate": 9.311667587615701e-06,
+      "loss": 1.4822916984558105,
+      "step": 222
+    },
+    {
+      "epoch": 0.8924302788844621,
+      "grad_norm": 0.8425886631011963,
+      "learning_rate": 9.29581491421405e-06,
+      "loss": 0.7406803965568542,
+      "step": 224
+    },
+    {
+      "epoch": 0.900398406374502,
+      "grad_norm": 0.30999937653541565,
+      "learning_rate": 9.279797243302453e-06,
+      "loss": 0.9925673604011536,
+      "step": 226
+    },
+    {
+      "epoch": 0.9083665338645418,
+      "grad_norm": 0.8192048072814941,
+      "learning_rate": 9.263615271141362e-06,
+      "loss": 1.0909143686294556,
+      "step": 228
+    },
+    {
+      "epoch": 0.9163346613545816,
+      "grad_norm": 1.8265670537948608,
+      "learning_rate": 9.247269701133125e-06,
+      "loss": 1.1720854043960571,
+      "step": 230
+    },
+    {
+      "epoch": 0.9243027888446215,
+      "grad_norm": 0.34351372718811035,
+      "learning_rate": 9.230761243791401e-06,
+      "loss": 1.327316403388977,
+      "step": 232
+    },
+    {
+      "epoch": 0.9322709163346613,
+      "grad_norm": 0.3349069356918335,
+      "learning_rate": 9.214090616710282e-06,
+      "loss": 0.7997048497200012,
+      "step": 234
+    },
+    {
+      "epoch": 0.9402390438247012,
+      "grad_norm": 0.3093612492084503,
+      "learning_rate": 9.197258544533098e-06,
+      "loss": 0.8246931433677673,
+      "step": 236
+    },
+    {
+      "epoch": 0.9482071713147411,
+      "grad_norm": 1.1106470823287964,
+      "learning_rate": 9.180265758920916e-06,
+      "loss": 0.48747754096984863,
+      "step": 238
+    },
+    {
+      "epoch": 0.9561752988047809,
+      "grad_norm": 1.0352427959442139,
+      "learning_rate": 9.163112998520741e-06,
+      "loss": 0.8584343791007996,
+      "step": 240
+    },
+    {
+      "epoch": 0.9641434262948207,
+      "grad_norm": 0.8169134259223938,
+      "learning_rate": 9.145801008933403e-06,
+      "loss": 1.1320377588272095,
+      "step": 242
+    },
+    {
+      "epoch": 0.9721115537848606,
+      "grad_norm": 0.30732667446136475,
+      "learning_rate": 9.128330542681154e-06,
+      "loss": 1.211287260055542,
+      "step": 244
+    },
+    {
+      "epoch": 0.9800796812749004,
+      "grad_norm": 0.3931141793727875,
+      "learning_rate": 9.110702359174944e-06,
+      "loss": 0.964515745639801,
+      "step": 246
+    },
+    {
+      "epoch": 0.9880478087649402,
+      "grad_norm": 0.24414241313934326,
+      "learning_rate": 9.092917224681427e-06,
+      "loss": 0.9973618388175964,
+      "step": 248
+    },
+    {
+      "epoch": 0.9960159362549801,
+      "grad_norm": 0.6048088073730469,
+      "learning_rate": 9.074975912289641e-06,
+      "loss": 1.0028462409973145,
+      "step": 250
+    },
+    {
+      "epoch": 1.00398406374502,
+      "grad_norm": 0.25815480947494507,
+      "learning_rate": 9.056879201877412e-06,
+      "loss": 0.9177858233451843,
+      "step": 252
+    },
+    {
+      "epoch": 1.0119521912350598,
+      "grad_norm": 0.2693774104118347,
+      "learning_rate": 9.038627880077441e-06,
+      "loss": 0.6799414157867432,
+      "step": 254
+    },
+    {
+      "epoch": 1.0199203187250996,
+      "grad_norm": 0.3611561059951782,
+      "learning_rate": 9.020222740243127e-06,
+      "loss": 1.1199322938919067,
+      "step": 256
+    },
+    {
+      "epoch": 1.0278884462151394,
+      "grad_norm": 0.3361198902130127,
+      "learning_rate": 9.001664582414066e-06,
+      "loss": 1.1035202741622925,
+      "step": 258
+    },
+    {
+      "epoch": 1.0358565737051793,
+      "grad_norm": 0.36350151896476746,
+      "learning_rate": 8.982954213281291e-06,
+      "loss": 1.0065547227859497,
+      "step": 260
+    },
+    {
+      "epoch": 1.043824701195219,
+      "grad_norm": 0.46773257851600647,
+      "learning_rate": 8.964092446152183e-06,
+      "loss": 0.9431231617927551,
+      "step": 262
+    },
+    {
+      "epoch": 1.051792828685259,
+      "grad_norm": 0.5551800727844238,
+      "learning_rate": 8.945080100915145e-06,
+      "loss": 1.034208059310913,
+      "step": 264
+    },
+    {
+      "epoch": 1.0597609561752988,
+      "grad_norm": 2.47458553314209,
+      "learning_rate": 8.925918004003941e-06,
+      "loss": 0.6088516116142273,
+      "step": 266
+    },
+    {
+      "epoch": 1.0677290836653386,
+      "grad_norm": 0.18113774061203003,
+      "learning_rate": 8.90660698836179e-06,
+      "loss": 0.8790088295936584,
+      "step": 268
+    },
+    {
+      "epoch": 1.0756972111553784,
+      "grad_norm": 0.2902234196662903,
+      "learning_rate": 8.887147893405145e-06,
+      "loss": 0.5367993116378784,
+      "step": 270
+    },
+    {
+      "epoch": 1.0836653386454183,
+      "grad_norm": 0.4603916108608246,
+      "learning_rate": 8.867541564987212e-06,
+      "loss": 1.2136603593826294,
+      "step": 272
+    },
+    {
+      "epoch": 1.091633466135458,
+      "grad_norm": 0.8974816799163818,
+      "learning_rate": 8.84778885536118e-06,
+      "loss": 0.5709693431854248,
+      "step": 274
+    },
+    {
+      "epoch": 1.099601593625498,
+      "grad_norm": 0.3268212378025055,
+      "learning_rate": 8.827890623143172e-06,
+      "loss": 0.8507805466651917,
+      "step": 276
+    },
+    {
+      "epoch": 1.1075697211155378,
+      "grad_norm": 0.280198872089386,
+      "learning_rate": 8.80784773327494e-06,
+      "loss": 0.7563758492469788,
+      "step": 278
+    },
+    {
+      "epoch": 1.1155378486055776,
+      "grad_norm": 0.3483583629131317,
+      "learning_rate": 8.787661056986236e-06,
+      "loss": 0.9553061127662659,
+      "step": 280
+    },
+    {
+      "epoch": 1.1235059760956174,
+      "grad_norm": 0.49651068449020386,
+      "learning_rate": 8.767331471756975e-06,
+      "loss": 1.0967178344726562,
+      "step": 282
+    },
+    {
+      "epoch": 1.1314741035856573,
+      "grad_norm": 0.6270463466644287,
+      "learning_rate": 8.746859861279068e-06,
+      "loss": 1.1502132415771484,
+      "step": 284
+    },
+    {
+      "epoch": 1.139442231075697,
+      "grad_norm": 0.7059192061424255,
+      "learning_rate": 8.726247115418017e-06,
+      "loss": 0.25774043798446655,
+      "step": 286
+    },
+    {
+      "epoch": 1.1474103585657371,
+      "grad_norm": 0.2589842677116394,
+      "learning_rate": 8.70549413017424e-06,
+      "loss": 0.7586540579795837,
+      "step": 288
+    },
+    {
+      "epoch": 1.155378486055777,
+      "grad_norm": 4.306430816650391,
+      "learning_rate": 8.684601807644114e-06,
+      "loss": 0.8887538313865662,
+      "step": 290
+    },
+    {
+      "epoch": 1.1633466135458168,
+      "grad_norm": 0.5187219381332397,
+      "learning_rate": 8.663571055980769e-06,
+      "loss": 0.4864007830619812,
+      "step": 292
+    },
+    {
+      "epoch": 1.1713147410358566,
+      "grad_norm": 0.26328423619270325,
+      "learning_rate": 8.642402789354609e-06,
+      "loss": 1.15906822681427,
+      "step": 294
+    },
+    {
+      "epoch": 1.1792828685258965,
+      "grad_norm": 0.4272761642932892,
+      "learning_rate": 8.62109792791358e-06,
+      "loss": 1.2134464979171753,
+      "step": 296
+    },
+    {
+      "epoch": 1.1872509960159363,
+      "grad_norm": 0.3709120452404022,
+      "learning_rate": 8.599657397743163e-06,
+      "loss": 1.276119589805603,
+      "step": 298
+    },
+    {
+      "epoch": 1.1952191235059761,
+      "grad_norm": 0.7744236588478088,
+      "learning_rate": 8.578082130826123e-06,
+      "loss": 0.5697095394134521,
+      "step": 300
+    },
+    {
+      "epoch": 1.203187250996016,
+      "grad_norm": 0.7287165522575378,
+      "learning_rate": 8.556373065002006e-06,
+      "loss": 0.8747298121452332,
+      "step": 302
+    },
+    {
+      "epoch": 1.2111553784860558,
+      "grad_norm": 0.5595633387565613,
+      "learning_rate": 8.534531143926359e-06,
+      "loss": 0.5382940769195557,
+      "step": 304
+    },
+    {
+      "epoch": 1.2191235059760956,
+      "grad_norm": 0.42395615577697754,
+      "learning_rate": 8.512557317029715e-06,
+      "loss": 1.1440248489379883,
+      "step": 306
+    },
+    {
+      "epoch": 1.2270916334661355,
+      "grad_norm": 1.672155737876892,
+      "learning_rate": 8.49045253947633e-06,
+      "loss": 0.9734455943107605,
+      "step": 308
+    },
+    {
+      "epoch": 1.2350597609561753,
+      "grad_norm": 0.4865654706954956,
+      "learning_rate": 8.46821777212266e-06,
+      "loss": 1.1162217855453491,
+      "step": 310
+    },
+    {
+      "epoch": 1.2430278884462151,
+      "grad_norm": 0.9238089323043823,
+      "learning_rate": 8.445853981475587e-06,
+      "loss": 1.0197659730911255,
+      "step": 312
+    },
+    {
+      "epoch": 1.250996015936255,
+      "grad_norm": 1.0905758142471313,
+      "learning_rate": 8.423362139650414e-06,
+      "loss": 0.8660633563995361,
+      "step": 314
+    },
+    {
+      "epoch": 1.2589641434262948,
+      "grad_norm": 0.5747548937797546,
+      "learning_rate": 8.40074322432861e-06,
+      "loss": 0.7378284335136414,
+      "step": 316
+    },
+    {
+      "epoch": 1.2669322709163346,
+      "grad_norm": 0.4243934452533722,
+      "learning_rate": 8.377998218715311e-06,
+      "loss": 0.6450012922286987,
+      "step": 318
+    },
+    {
+      "epoch": 1.2749003984063745,
+      "grad_norm": 0.8766657710075378,
+      "learning_rate": 8.355128111496569e-06,
+      "loss": 0.4933542013168335,
+      "step": 320
+    },
+    {
+      "epoch": 1.2828685258964143,
+      "grad_norm": 0.43255963921546936,
+      "learning_rate": 8.332133896796399e-06,
+      "loss": 0.969576895236969,
+      "step": 322
+    },
+    {
+      "epoch": 1.2908366533864541,
+      "grad_norm": 1.0938221216201782,
+      "learning_rate": 8.309016574133545e-06,
+      "loss": 0.8896313309669495,
+      "step": 324
+    },
+    {
+      "epoch": 1.298804780876494,
+      "grad_norm": 1.1368612051010132,
+      "learning_rate": 8.28577714837805e-06,
+      "loss": 1.2510038614273071,
+      "step": 326
+    },
+    {
+      "epoch": 1.3067729083665338,
+      "grad_norm": 0.2708609104156494,
+      "learning_rate": 8.262416629707555e-06,
+      "loss": 0.7044373154640198,
+      "step": 328
+    },
+    {
+      "epoch": 1.3147410358565736,
+      "grad_norm": 0.3936620056629181,
+      "learning_rate": 8.238936033563415e-06,
+      "loss": 0.8204818367958069,
+      "step": 330
+    },
+    {
+      "epoch": 1.3227091633466135,
+      "grad_norm": 1.3723121881484985,
+      "learning_rate": 8.215336380606533e-06,
+      "loss": 0.9094972014427185,
+      "step": 332
+    },
+    {
+      "epoch": 1.3306772908366533,
+      "grad_norm": 0.7730778455734253,
+      "learning_rate": 8.191618696673012e-06,
+      "loss": 0.6840326189994812,
+      "step": 334
+    },
+    {
+      "epoch": 1.3386454183266931,
+      "grad_norm": 0.8889527320861816,
+      "learning_rate": 8.167784012729555e-06,
+      "loss": 0.8412512540817261,
+      "step": 336
+    },
+    {
+      "epoch": 1.3466135458167332,
+      "grad_norm": 0.5860934257507324,
+      "learning_rate": 8.143833364828654e-06,
+      "loss": 1.1132867336273193,
+      "step": 338
+    },
+    {
+      "epoch": 1.354581673306773,
+      "grad_norm": 0.3146251440048218,
+      "learning_rate": 8.119767794063553e-06,
+      "loss": 1.087275505065918,
+      "step": 340
+    },
+    {
+      "epoch": 1.3625498007968129,
+      "grad_norm": 0.3383582830429077,
+      "learning_rate": 8.095588346523e-06,
+      "loss": 1.1487057209014893,
+      "step": 342
+    },
+    {
+      "epoch": 1.3705179282868527,
+      "grad_norm": 0.3725880980491638,
+      "learning_rate": 8.071296073245759e-06,
+      "loss": 1.101293921470642,
+      "step": 344
+    },
+    {
+      "epoch": 1.3784860557768925,
+      "grad_norm": 0.47705674171447754,
+      "learning_rate": 8.046892030174938e-06,
+      "loss": 0.7032272219657898,
+      "step": 346
+    },
+    {
+      "epoch": 1.3864541832669324,
+      "grad_norm": 0.40366417169570923,
+      "learning_rate": 8.022377278112089e-06,
+      "loss": 1.1423650979995728,
+      "step": 348
+    },
+    {
+      "epoch": 1.3944223107569722,
+      "grad_norm": 0.6584059000015259,
+      "learning_rate": 7.997752882671088e-06,
+      "loss": 0.6649424433708191,
+      "step": 350
+    },
+    {
+      "epoch": 1.402390438247012,
+      "grad_norm": 0.6367621421813965,
+      "learning_rate": 7.97301991423182e-06,
+      "loss": 1.1485172510147095,
+      "step": 352
+    },
+    {
+      "epoch": 1.4103585657370519,
+      "grad_norm": 0.6186166405677795,
+      "learning_rate": 7.948179447893652e-06,
+      "loss": 1.0935237407684326,
+      "step": 354
+    },
+    {
+      "epoch": 1.4183266932270917,
+      "grad_norm": 0.4727344810962677,
+      "learning_rate": 7.923232563428693e-06,
+      "loss": 0.8808469772338867,
+      "step": 356
+    },
+    {
+      "epoch": 1.4262948207171315,
+      "grad_norm": 0.4122134745121002,
+      "learning_rate": 7.898180345234875e-06,
+      "loss": 1.1339887380599976,
+      "step": 358
+    },
+    {
+      "epoch": 1.4342629482071714,
+      "grad_norm": 0.9194333553314209,
+      "learning_rate": 7.873023882288793e-06,
+      "loss": 0.5680944323539734,
+      "step": 360
+    },
+    {
+      "epoch": 1.4422310756972112,
+      "grad_norm": 0.3512038588523865,
+      "learning_rate": 7.847764268098387e-06,
+      "loss": 0.665905773639679,
+      "step": 362
+    },
+    {
+      "epoch": 1.450199203187251,
+      "grad_norm": 0.7209855914115906,
+      "learning_rate": 7.822402600655405e-06,
+      "loss": 1.0396764278411865,
+      "step": 364
+    },
+    {
+      "epoch": 1.4581673306772909,
+      "grad_norm": 0.43510812520980835,
+      "learning_rate": 7.79693998238767e-06,
+      "loss": 1.083251953125,
+      "step": 366
+    },
+    {
+      "epoch": 1.4661354581673307,
+      "grad_norm": 0.32131174206733704,
+      "learning_rate": 7.771377520111163e-06,
+      "loss": 0.589643120765686,
+      "step": 368
+    },
+    {
+      "epoch": 1.4741035856573705,
+      "grad_norm": 1.8862578868865967,
+      "learning_rate": 7.745716324981912e-06,
+      "loss": 0.6705837249755859,
+      "step": 370
+    },
+    {
+      "epoch": 1.4820717131474104,
+      "grad_norm": 0.6846002340316772,
+      "learning_rate": 7.71995751244769e-06,
+      "loss": 1.2591749429702759,
+      "step": 372
+    },
+    {
+      "epoch": 1.4900398406374502,
+      "grad_norm": 0.3747018873691559,
+      "learning_rate": 7.694102202199536e-06,
+      "loss": 1.0850481986999512,
+      "step": 374
+    },
+    {
+      "epoch": 1.49800796812749,
+      "grad_norm": 0.24257667362689972,
+      "learning_rate": 7.66815151812307e-06,
+      "loss": 1.0923634767532349,
+      "step": 376
+    },
+    {
+      "epoch": 1.5059760956175299,
+      "grad_norm": 0.4980308711528778,
+      "learning_rate": 7.642106588249651e-06,
+      "loss": 0.8712183237075806,
+      "step": 378
+    },
+    {
+      "epoch": 1.5139442231075697,
+      "grad_norm": 0.2823687791824341,
+      "learning_rate": 7.6159685447073376e-06,
+      "loss": 1.1045010089874268,
+      "step": 380
+    },
+    {
+      "epoch": 1.5219123505976095,
+      "grad_norm": 0.25547516345977783,
+      "learning_rate": 7.589738523671685e-06,
+      "loss": 0.6853191256523132,
+      "step": 382
+    },
+    {
+      "epoch": 1.5298804780876494,
+      "grad_norm": 0.3622785806655884,
+      "learning_rate": 7.563417665316338e-06,
+      "loss": 0.7108783721923828,
+      "step": 384
+    },
+    {
+      "epoch": 1.5378486055776892,
+      "grad_norm": 1.2122690677642822,
+      "learning_rate": 7.5370071137634984e-06,
+      "loss": 0.9623269438743591,
+      "step": 386
+    },
+    {
+      "epoch": 1.545816733067729,
+      "grad_norm": 0.3348121643066406,
+      "learning_rate": 7.510508017034165e-06,
+      "loss": 1.1149513721466064,
+      "step": 388
+    },
+    {
+      "epoch": 1.5537848605577689,
+      "grad_norm": 0.9548903703689575,
+      "learning_rate": 7.483921526998247e-06,
+      "loss": 0.5825637578964233,
+      "step": 390
+    },
+    {
+      "epoch": 1.5617529880478087,
+      "grad_norm": 0.3586859703063965,
+      "learning_rate": 7.4572487993244865e-06,
+      "loss": 0.5829859972000122,
+      "step": 392
+    },
+    {
+      "epoch": 1.5697211155378485,
+      "grad_norm": 0.7318173050880432,
+      "learning_rate": 7.430490993430232e-06,
+      "loss": 1.035295009613037,
+      "step": 394
+    },
+    {
+      "epoch": 1.5776892430278884,
+      "grad_norm": 0.366339772939682,
+      "learning_rate": 7.403649272431027e-06,
+      "loss": 0.9984535574913025,
+      "step": 396
+    },
+    {
+      "epoch": 1.5856573705179282,
+      "grad_norm": 0.1879645735025406,
+      "learning_rate": 7.376724803090069e-06,
+      "loss": 1.1060404777526855,
+      "step": 398
+    },
+    {
+      "epoch": 1.593625498007968,
+      "grad_norm": 0.41772928833961487,
+      "learning_rate": 7.349718755767473e-06,
+      "loss": 0.592603325843811,
+      "step": 400
+    },
+    {
+      "epoch": 1.6015936254980079,
+      "grad_norm": 0.5612187385559082,
+      "learning_rate": 7.322632304369416e-06,
+      "loss": 1.0864520072937012,
+      "step": 402
+    },
+    {
+      "epoch": 1.6095617529880477,
+      "grad_norm": 0.34493568539619446,
+      "learning_rate": 7.2954666262970965e-06,
+      "loss": 1.2245590686798096,
+      "step": 404
+    },
+    {
+      "epoch": 1.6175298804780875,
+      "grad_norm": 0.3961571156978607,
+      "learning_rate": 7.268222902395562e-06,
+      "loss": 0.7078885436058044,
+      "step": 406
+    },
+    {
+      "epoch": 1.6254980079681274,
+      "grad_norm": 0.2080133855342865,
+      "learning_rate": 7.240902316902373e-06,
+      "loss": 0.5460065603256226,
+      "step": 408
+    },
+    {
+      "epoch": 1.6334661354581672,
+      "grad_norm": 0.23523752391338348,
+      "learning_rate": 7.213506057396134e-06,
+      "loss": 0.7281398177146912,
+      "step": 410
+    },
+    {
+      "epoch": 1.641434262948207,
+      "grad_norm": 0.6262185573577881,
+      "learning_rate": 7.186035314744872e-06,
+      "loss": 0.4776509404182434,
+      "step": 412
+    },
+    {
+      "epoch": 1.6494023904382469,
+      "grad_norm": 0.25708886981010437,
+      "learning_rate": 7.1584912830542575e-06,
+      "loss": 0.7877987623214722,
+      "step": 414
+    },
+    {
+      "epoch": 1.6573705179282867,
+      "grad_norm": 0.3205985724925995,
+      "learning_rate": 7.130875159615716e-06,
+      "loss": 0.7444572448730469,
+      "step": 416
+    },
+    {
+      "epoch": 1.6653386454183265,
+      "grad_norm": 0.6451170444488525,
+      "learning_rate": 7.103188144854376e-06,
+      "loss": 0.9179713129997253,
+      "step": 418
+    },
+    {
+      "epoch": 1.6733067729083664,
+      "grad_norm": 0.33163195848464966,
+      "learning_rate": 7.075431442276888e-06,
+      "loss": 1.1229826211929321,
+      "step": 420
+    },
+    {
+      "epoch": 1.6812749003984062,
+      "grad_norm": 0.32244348526000977,
+      "learning_rate": 7.047606258419108e-06,
+      "loss": 0.9841935038566589,
+      "step": 422
+    },
+    {
+      "epoch": 1.6892430278884463,
+      "grad_norm": 0.43349865078926086,
+      "learning_rate": 7.019713802793663e-06,
+      "loss": 0.6049963235855103,
+      "step": 424
+    },
+    {
+      "epoch": 1.697211155378486,
+      "grad_norm": 0.2760488986968994,
+      "learning_rate": 6.99175528783736e-06,
+      "loss": 1.1549838781356812,
+      "step": 426
+    },
+    {
+      "epoch": 1.705179282868526,
+      "grad_norm": 0.27986299991607666,
+      "learning_rate": 6.963731928858495e-06,
+      "loss": 1.0964317321777344,
+      "step": 428
+    },
+    {
+      "epoch": 1.7131474103585658,
+      "grad_norm": 0.2775329053401947,
+      "learning_rate": 6.935644943984017e-06,
+      "loss": 0.6988530158996582,
+      "step": 430
+    },
+    {
+      "epoch": 1.7211155378486056,
+      "grad_norm": 0.8730498552322388,
+      "learning_rate": 6.9074955541065865e-06,
+      "loss": 0.8607664108276367,
+      "step": 432
+    },
+    {
+      "epoch": 1.7290836653386454,
+      "grad_norm": 0.5013867616653442,
+      "learning_rate": 6.8792849828315054e-06,
+      "loss": 0.6253218650817871,
+      "step": 434
+    },
+    {
+      "epoch": 1.7370517928286853,
+      "grad_norm": 0.85249263048172,
+      "learning_rate": 6.851014456423515e-06,
+      "loss": 0.39246758818626404,
+      "step": 436
+    },
+    {
+      "epoch": 1.745019920318725,
+      "grad_norm": 0.3706222176551819,
+      "learning_rate": 6.822685203753508e-06,
+      "loss": 0.73777836561203,
+      "step": 438
+    },
+    {
+      "epoch": 1.752988047808765,
+      "grad_norm": 1.949631929397583,
+      "learning_rate": 6.794298456245111e-06,
+      "loss": 0.7235686182975769,
+      "step": 440
+    },
+    {
+      "epoch": 1.7609561752988048,
+      "grad_norm": 0.6743807792663574,
+      "learning_rate": 6.765855447821144e-06,
+      "loss": 1.4941140413284302,
+      "step": 442
+    },
+    {
+      "epoch": 1.7689243027888446,
+      "grad_norm": 0.45370909571647644,
+      "learning_rate": 6.737357414849993e-06,
+      "loss": 0.7234033346176147,
+      "step": 444
+    },
+    {
+      "epoch": 1.7768924302788844,
+      "grad_norm": 0.23447301983833313,
+      "learning_rate": 6.70880559609187e-06,
+      "loss": 1.0368897914886475,
+      "step": 446
+    },
+    {
+      "epoch": 1.7848605577689243,
+      "grad_norm": 0.5336112976074219,
+      "learning_rate": 6.680201232644958e-06,
+      "loss": 1.0760776996612549,
+      "step": 448
+    },
+    {
+      "epoch": 1.792828685258964,
+      "grad_norm": 0.20120948553085327,
+      "learning_rate": 6.651545567891471e-06,
+      "loss": 0.821214497089386,
+      "step": 450
+    },
+    {
+      "epoch": 1.800796812749004,
+      "grad_norm": 0.23334261775016785,
+      "learning_rate": 6.622839847443596e-06,
+      "loss": 0.49071204662323,
+      "step": 452
+    },
+    {
+      "epoch": 1.8087649402390438,
+      "grad_norm": 0.940449595451355,
+      "learning_rate": 6.59408531908936e-06,
+      "loss": 0.9011412858963013,
+      "step": 454
+    },
+    {
+      "epoch": 1.8167330677290838,
+      "grad_norm": 1.4027504920959473,
+      "learning_rate": 6.565283232738386e-06,
+      "loss": 0.6551702618598938,
+      "step": 456
+    },
+    {
+      "epoch": 1.8247011952191237,
+      "grad_norm": 0.38345828652381897,
+      "learning_rate": 6.536434840367555e-06,
+      "loss": 1.022027850151062,
+      "step": 458
+    },
+    {
+      "epoch": 1.8326693227091635,
+      "grad_norm": 0.40730491280555725,
+      "learning_rate": 6.507541395966591e-06,
+      "loss": 0.7417793273925781,
+      "step": 460
+    },
+    {
+      "epoch": 1.8406374501992033,
+      "grad_norm": 0.27808839082717896,
+      "learning_rate": 6.47860415548356e-06,
+      "loss": 0.8030527234077454,
+      "step": 462
+    },
+    {
+      "epoch": 1.8486055776892432,
+      "grad_norm": 0.20615167915821075,
+      "learning_rate": 6.449624376770259e-06,
+      "loss": 0.123322032392025,
+      "step": 464
+    },
+    {
+      "epoch": 1.856573705179283,
+      "grad_norm": 0.45746463537216187,
+      "learning_rate": 6.420603319527554e-06,
+      "loss": 1.1299569606781006,
+      "step": 466
+    },
+    {
+      "epoch": 1.8645418326693228,
+      "grad_norm": 0.3837216794490814,
+      "learning_rate": 6.3915422452506146e-06,
+      "loss": 0.5265716314315796,
+      "step": 468
+    },
+    {
+      "epoch": 1.8725099601593627,
+      "grad_norm": 0.37781667709350586,
+      "learning_rate": 6.362442417174082e-06,
+      "loss": 1.1209043264389038,
+      "step": 470
+    },
+    {
+      "epoch": 1.8804780876494025,
+      "grad_norm": 3.5912818908691406,
+      "learning_rate": 6.33330510021716e-06,
+      "loss": 0.9933595657348633,
+      "step": 472
+    },
+    {
+      "epoch": 1.8884462151394423,
+      "grad_norm": 0.7557405233383179,
+      "learning_rate": 6.304131560928622e-06,
+      "loss": 1.083042025566101,
+      "step": 474
+    },
+    {
+      "epoch": 1.8964143426294822,
+      "grad_norm": 0.24848543107509613,
+      "learning_rate": 6.274923067431775e-06,
+      "loss": 0.9510005712509155,
+      "step": 476
+    },
+    {
+      "epoch": 1.904382470119522,
+      "grad_norm": 0.27726584672927856,
+      "learning_rate": 6.2456808893693155e-06,
+      "loss": 0.8397172689437866,
+      "step": 478
+    },
+    {
+      "epoch": 1.9123505976095618,
+      "grad_norm": 0.504697859287262,
+      "learning_rate": 6.216406297848155e-06,
+      "loss": 0.6099559664726257,
+      "step": 480
+    },
+    {
+      "epoch": 1.9203187250996017,
+      "grad_norm": 0.6044100522994995,
+      "learning_rate": 6.187100565384162e-06,
+      "loss": 0.7560096979141235,
+      "step": 482
+    },
+    {
+      "epoch": 1.9282868525896415,
+      "grad_norm": 0.5274583697319031,
+      "learning_rate": 6.157764965846841e-06,
+      "loss": 1.0795762538909912,
+      "step": 484
+    },
+    {
+      "epoch": 1.9362549800796813,
+      "grad_norm": 0.34211012721061707,
+      "learning_rate": 6.128400774403974e-06,
+      "loss": 1.0622626543045044,
+      "step": 486
+    },
+    {
+      "epoch": 1.9442231075697212,
+      "grad_norm": 0.24703830480575562,
+      "learning_rate": 6.0990092674661834e-06,
+      "loss": 1.069478154182434,
+      "step": 488
+    },
+    {
+      "epoch": 1.952191235059761,
+      "grad_norm": 0.23846203088760376,
+      "learning_rate": 6.069591722631446e-06,
+      "loss": 1.0753988027572632,
+      "step": 490
+    },
+    {
+      "epoch": 1.9601593625498008,
+      "grad_norm": 0.31255602836608887,
+      "learning_rate": 6.040149418629562e-06,
+      "loss": 0.7481006979942322,
+      "step": 492
+    },
+    {
+      "epoch": 1.9681274900398407,
+      "grad_norm": 3.418362617492676,
+      "learning_rate": 6.010683635266569e-06,
+      "loss": 0.6890715956687927,
+      "step": 494
+    },
+    {
+      "epoch": 1.9760956175298805,
+      "grad_norm": 0.2767847180366516,
+      "learning_rate": 5.981195653369113e-06,
+      "loss": 1.0435179471969604,
+      "step": 496
+    },
+    {
+      "epoch": 1.9840637450199203,
+      "grad_norm": 0.8114778399467468,
+      "learning_rate": 5.95168675472877e-06,
+      "loss": 0.8171101808547974,
+      "step": 498
+    },
+    {
+      "epoch": 1.9920318725099602,
+      "grad_norm": 0.3478597402572632,
+      "learning_rate": 5.922158222046334e-06,
+      "loss": 1.152536153793335,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.5647305846214294,
+      "learning_rate": 5.892611338876053e-06,
+      "loss": 0.6063491702079773,
+      "step": 502
+    },
+    {
+      "epoch": 2.00796812749004,
+      "grad_norm": 0.28193220496177673,
+      "learning_rate": 5.8630473895698426e-06,
+      "loss": 0.5733928084373474,
+      "step": 504
+    },
+    {
+      "epoch": 2.0159362549800797,
+      "grad_norm": 0.5333064198493958,
+      "learning_rate": 5.833467659221452e-06,
+      "loss": 0.5601115822792053,
+      "step": 506
+    },
+    {
+      "epoch": 2.0239043824701195,
+      "grad_norm": 0.09115895628929138,
+      "learning_rate": 5.803873433610604e-06,
+      "loss": 0.5059247612953186,
+      "step": 508
+    },
+    {
+      "epoch": 2.0318725099601593,
+      "grad_norm": 0.3979955315589905,
+      "learning_rate": 5.774265999147108e-06,
+      "loss": 0.7920820116996765,
+      "step": 510
+    },
+    {
+      "epoch": 2.039840637450199,
+      "grad_norm": 0.3301240801811218,
+      "learning_rate": 5.744646642814935e-06,
+      "loss": 0.6867862939834595,
+      "step": 512
+    },
+    {
+      "epoch": 2.047808764940239,
+      "grad_norm": 0.834158182144165,
+      "learning_rate": 5.715016652116284e-06,
+      "loss": 0.341141015291214,
+      "step": 514
+    },
+    {
+      "epoch": 2.055776892430279,
+      "grad_norm": 0.3339720368385315,
+      "learning_rate": 5.68537731501561e-06,
+      "loss": 0.8390861749649048,
+      "step": 516
+    },
+    {
+      "epoch": 2.0637450199203187,
+      "grad_norm": 0.6279764175415039,
+      "learning_rate": 5.655729919883639e-06,
+      "loss": 0.5062633156776428,
+      "step": 518
+    },
+    {
+      "epoch": 2.0717131474103585,
+      "grad_norm": 0.5732558369636536,
+      "learning_rate": 5.626075755441366e-06,
+      "loss": 0.7195804715156555,
+      "step": 520
+    },
+    {
+      "epoch": 2.0796812749003983,
+      "grad_norm": 0.9404535889625549,
+      "learning_rate": 5.596416110704034e-06,
+      "loss": 0.9061080813407898,
+      "step": 522
+    },
+    {
+      "epoch": 2.087649402390438,
+      "grad_norm": 0.6268462538719177,
+      "learning_rate": 5.566752274925109e-06,
+      "loss": 0.8412163257598877,
+      "step": 524
+    },
+    {
+      "epoch": 2.095617529880478,
+      "grad_norm": 0.7196136116981506,
+      "learning_rate": 5.5370855375402345e-06,
+      "loss": 0.41694265604019165,
+      "step": 526
+    },
+    {
+      "epoch": 2.103585657370518,
+      "grad_norm": 0.5400736331939697,
+      "learning_rate": 5.507417188111176e-06,
+      "loss": 0.6703310012817383,
+      "step": 528
+    },
+    {
+      "epoch": 2.1115537848605577,
+      "grad_norm": 0.1484118551015854,
+      "learning_rate": 5.477748516269779e-06,
+      "loss": 0.5773021578788757,
+      "step": 530
+    },
+    {
+      "epoch": 2.1195219123505975,
+      "grad_norm": 0.21181708574295044,
+      "learning_rate": 5.448080811661897e-06,
+      "loss": 0.19861428439617157,
+      "step": 532
+    },
+    {
+      "epoch": 2.1274900398406373,
+      "grad_norm": 0.4062117338180542,
+      "learning_rate": 5.418415363891346e-06,
+      "loss": 0.682986855506897,
+      "step": 534
+    },
+    {
+      "epoch": 2.135458167330677,
+      "grad_norm": 1.2481952905654907,
+      "learning_rate": 5.3887534624638335e-06,
+      "loss": 0.6050370931625366,
+      "step": 536
+    },
+    {
+      "epoch": 2.143426294820717,
+      "grad_norm": 0.8576554656028748,
+      "learning_rate": 5.359096396730924e-06,
+      "loss": 0.6381558775901794,
+      "step": 538
+    },
+    {
+      "epoch": 2.151394422310757,
+      "grad_norm": 0.32604241371154785,
+      "learning_rate": 5.329445455833975e-06,
+      "loss": 0.8527320623397827,
+      "step": 540
+    },
+    {
+      "epoch": 2.1593625498007967,
+      "grad_norm": 1.2591558694839478,
+      "learning_rate": 5.299801928648105e-06,
+      "loss": 0.8114980459213257,
+      "step": 542
+    },
+    {
+      "epoch": 2.1673306772908365,
+      "grad_norm": 0.2215310037136078,
+      "learning_rate": 5.270167103726181e-06,
+      "loss": 0.6226505637168884,
+      "step": 544
+    },
+    {
+      "epoch": 2.1752988047808763,
+      "grad_norm": 0.5441330671310425,
+      "learning_rate": 5.240542269242794e-06,
+      "loss": 0.602466881275177,
+      "step": 546
+    },
+    {
+      "epoch": 2.183266932270916,
+      "grad_norm": 0.3258216977119446,
+      "learning_rate": 5.210928712938262e-06,
+      "loss": 0.8800083994865417,
+      "step": 548
+    },
+    {
+      "epoch": 2.191235059760956,
+      "grad_norm": 0.27899324893951416,
+      "learning_rate": 5.181327722062667e-06,
+      "loss": 0.94955974817276,
+      "step": 550
+    },
+    {
+      "epoch": 2.199203187250996,
+      "grad_norm": 0.3722408413887024,
+      "learning_rate": 5.151740583319892e-06,
+      "loss": 0.5929532647132874,
+      "step": 552
+    },
+    {
+      "epoch": 2.2071713147410357,
+      "grad_norm": 0.7487623691558838,
+      "learning_rate": 5.122168582811689e-06,
+      "loss": 0.4511300027370453,
+      "step": 554
+    },
+    {
+      "epoch": 2.2151394422310755,
+      "grad_norm": 0.5579084157943726,
+      "learning_rate": 5.092613005981781e-06,
+      "loss": 0.616023063659668,
+      "step": 556
+    },
+    {
+      "epoch": 2.2231075697211153,
+      "grad_norm": 0.5267292261123657,
+      "learning_rate": 5.063075137559979e-06,
+      "loss": 0.8207051753997803,
+      "step": 558
+    },
+    {
+      "epoch": 2.231075697211155,
+      "grad_norm": 0.8398407697677612,
+      "learning_rate": 5.033556261506343e-06,
+      "loss": 0.8526216149330139,
+      "step": 560
+    },
+    {
+      "epoch": 2.239043824701195,
+      "grad_norm": 0.2584734857082367,
+      "learning_rate": 5.0040576609553645e-06,
+      "loss": 0.9668092727661133,
+      "step": 562
+    },
+    {
+      "epoch": 2.247011952191235,
+      "grad_norm": 1.4895679950714111,
+      "learning_rate": 4.974580618160194e-06,
+      "loss": 0.6944220066070557,
+      "step": 564
+    },
+    {
+      "epoch": 2.2549800796812747,
+      "grad_norm": 0.27689334750175476,
+      "learning_rate": 4.945126414436908e-06,
+      "loss": 0.5737918019294739,
+      "step": 566
+    },
+    {
+      "epoch": 2.2629482071713145,
+      "grad_norm": 0.3451761305332184,
+      "learning_rate": 4.915696330108804e-06,
+      "loss": 0.7355297803878784,
+      "step": 568
+    },
+    {
+      "epoch": 2.2709163346613543,
+      "grad_norm": 0.38218218088150024,
+      "learning_rate": 4.886291644450751e-06,
+      "loss": 0.9852412939071655,
+      "step": 570
+    },
+    {
+      "epoch": 2.278884462151394,
+      "grad_norm": 3.1000728607177734,
+      "learning_rate": 4.8569136356335845e-06,
+      "loss": 0.5935670137405396,
+      "step": 572
+    },
+    {
+      "epoch": 2.2868525896414345,
+      "grad_norm": 0.3808981478214264,
+      "learning_rate": 4.827563580668544e-06,
+      "loss": 0.6737416386604309,
+      "step": 574
+    },
+    {
+      "epoch": 2.2948207171314743,
+      "grad_norm": 0.2438511699438095,
+      "learning_rate": 4.798242755351755e-06,
+      "loss": 0.23764614760875702,
+      "step": 576
+    },
+    {
+      "epoch": 2.302788844621514,
+      "grad_norm": 0.5194549560546875,
+      "learning_rate": 4.768952434208795e-06,
+      "loss": 0.5982046723365784,
+      "step": 578
+    },
+    {
+      "epoch": 2.310756972111554,
+      "grad_norm": 0.6223345398902893,
+      "learning_rate": 4.7396938904392656e-06,
+      "loss": 0.5968120098114014,
+      "step": 580
+    },
+    {
+      "epoch": 2.318725099601594,
+      "grad_norm": 0.3958941698074341,
+      "learning_rate": 4.71046839586147e-06,
+      "loss": 0.5722007155418396,
+      "step": 582
+    },
+    {
+      "epoch": 2.3266932270916336,
+      "grad_norm": 0.2528054118156433,
+      "learning_rate": 4.681277220857108e-06,
+      "loss": 0.7101739645004272,
+      "step": 584
+    },
+    {
+      "epoch": 2.3346613545816735,
+      "grad_norm": 0.3789544701576233,
+      "learning_rate": 4.6521216343160824e-06,
+      "loss": 0.6098624467849731,
+      "step": 586
+    },
+    {
+      "epoch": 2.3426294820717133,
+      "grad_norm": 0.850007951259613,
+      "learning_rate": 4.623002903581316e-06,
+      "loss": 0.7937362194061279,
+      "step": 588
+    },
+    {
+      "epoch": 2.350597609561753,
+      "grad_norm": 0.2743369936943054,
+      "learning_rate": 4.593922294393677e-06,
+      "loss": 0.9130489826202393,
+      "step": 590
+    },
+    {
+      "epoch": 2.358565737051793,
+      "grad_norm": 0.35594576597213745,
+      "learning_rate": 4.564881070836955e-06,
+      "loss": 0.9428669214248657,
+      "step": 592
+    },
+    {
+      "epoch": 2.366533864541833,
+      "grad_norm": 0.38490182161331177,
+      "learning_rate": 4.535880495282916e-06,
+      "loss": 0.627445638179779,
+      "step": 594
+    },
+    {
+      "epoch": 2.3745019920318726,
+      "grad_norm": 0.4483906030654907,
+      "learning_rate": 4.506921828336427e-06,
+      "loss": 0.8550093173980713,
+      "step": 596
+    },
+    {
+      "epoch": 2.3824701195219125,
+      "grad_norm": 0.22079698741436005,
+      "learning_rate": 4.478006328780657e-06,
+      "loss": 0.6375987529754639,
+      "step": 598
+    },
+    {
+      "epoch": 2.3904382470119523,
+      "grad_norm": 1.2619231939315796,
+      "learning_rate": 4.44913525352237e-06,
+      "loss": 0.9383031129837036,
+      "step": 600
+    },
+    {
+      "epoch": 2.398406374501992,
+      "grad_norm": 0.256187379360199,
+      "learning_rate": 4.420309857537277e-06,
+      "loss": 0.9279384016990662,
+      "step": 602
+    },
+    {
+      "epoch": 2.406374501992032,
+      "grad_norm": 0.29454106092453003,
+      "learning_rate": 4.39153139381549e-06,
+      "loss": 0.5474885702133179,
+      "step": 604
+    },
+    {
+      "epoch": 2.414342629482072,
+      "grad_norm": 0.8071082830429077,
+      "learning_rate": 4.3628011133070605e-06,
+      "loss": 0.22221192717552185,
+      "step": 606
+    },
+    {
+      "epoch": 2.4223107569721116,
+      "grad_norm": 0.1772623509168625,
+      "learning_rate": 4.334120264867596e-06,
+      "loss": 0.43757298588752747,
+      "step": 608
+    },
+    {
+      "epoch": 2.4302788844621515,
+      "grad_norm": 0.15431374311447144,
+      "learning_rate": 4.305490095203979e-06,
+      "loss": 0.7972186803817749,
+      "step": 610
+    },
+    {
+      "epoch": 2.4382470119521913,
+      "grad_norm": 0.6691494584083557,
+      "learning_rate": 4.27691184882017e-06,
+      "loss": 0.6454163193702698,
+      "step": 612
+    },
+    {
+      "epoch": 2.446215139442231,
+      "grad_norm": 0.32318800687789917,
+      "learning_rate": 4.248386767963121e-06,
+      "loss": 0.5316993594169617,
+      "step": 614
+    },
+    {
+      "epoch": 2.454183266932271,
+      "grad_norm": 0.33901235461235046,
+      "learning_rate": 4.219916092568764e-06,
+      "loss": 0.8095099329948425,
+      "step": 616
+    },
+    {
+      "epoch": 2.462151394422311,
+      "grad_norm": 0.7935246825218201,
+      "learning_rate": 4.191501060208121e-06,
+      "loss": 0.636178731918335,
+      "step": 618
+    },
+    {
+      "epoch": 2.4701195219123506,
+      "grad_norm": 0.5776334404945374,
+      "learning_rate": 4.163142906033512e-06,
+      "loss": 0.7154951691627502,
+      "step": 620
+    },
+    {
+      "epoch": 2.4780876494023905,
+      "grad_norm": 0.4138113856315613,
+      "learning_rate": 4.134842862724854e-06,
+      "loss": 0.7439125180244446,
+      "step": 622
+    },
+    {
+      "epoch": 2.4860557768924303,
+      "grad_norm": 0.4851977825164795,
+      "learning_rate": 4.106602160436088e-06,
+      "loss": 0.7215173840522766,
+      "step": 624
+    },
+    {
+      "epoch": 2.49402390438247,
+      "grad_norm": 0.3580200970172882,
+      "learning_rate": 4.078422026741706e-06,
+      "loss": 0.5086095333099365,
+      "step": 626
+    },
+    {
+      "epoch": 2.50199203187251,
+      "grad_norm": 0.8635292053222656,
+      "learning_rate": 4.050303686583386e-06,
+      "loss": 0.9134188294410706,
+      "step": 628
+    },
+    {
+      "epoch": 2.50996015936255,
+      "grad_norm": 0.4788064956665039,
+      "learning_rate": 4.0222483622167465e-06,
+      "loss": 0.4058358371257782,
+      "step": 630
+    },
+    {
+      "epoch": 2.5179282868525896,
+      "grad_norm": 0.9498984217643738,
+      "learning_rate": 3.9942572731582185e-06,
+      "loss": 0.4294551908969879,
+      "step": 632
+    },
+    {
+      "epoch": 2.5258964143426295,
+      "grad_norm": 0.2861036956310272,
+      "learning_rate": 3.966331636132039e-06,
+      "loss": 0.9240294694900513,
+      "step": 634
+    },
+    {
+      "epoch": 2.5338645418326693,
+      "grad_norm": 0.3773558735847473,
+      "learning_rate": 3.938472665017355e-06,
+      "loss": 0.8665947914123535,
+      "step": 636
+    },
+    {
+      "epoch": 2.541832669322709,
+      "grad_norm": 0.518242359161377,
+      "learning_rate": 3.91068157079546e-06,
+      "loss": 0.9016304612159729,
+      "step": 638
+    },
+    {
+      "epoch": 2.549800796812749,
+      "grad_norm": 0.20046387612819672,
+      "learning_rate": 3.882959561497159e-06,
+      "loss": 0.5296582579612732,
+      "step": 640
+    },
+    {
+      "epoch": 2.557768924302789,
+      "grad_norm": 0.6685121059417725,
+      "learning_rate": 3.855307842150259e-06,
+      "loss": 0.8770003914833069,
+      "step": 642
+    },
+    {
+      "epoch": 2.5657370517928286,
+      "grad_norm": 0.33204641938209534,
+      "learning_rate": 3.827727614727177e-06,
+      "loss": 0.8915793299674988,
+      "step": 644
+    },
+    {
+      "epoch": 2.5737051792828685,
+      "grad_norm": 0.4422305226325989,
+      "learning_rate": 3.8002200780927e-06,
+      "loss": 0.5884954929351807,
+      "step": 646
+    },
+    {
+      "epoch": 2.5816733067729083,
+      "grad_norm": 0.8002772927284241,
+      "learning_rate": 3.772786427951878e-06,
+      "loss": 0.9170203804969788,
+      "step": 648
+    },
+    {
+      "epoch": 2.589641434262948,
+      "grad_norm": 0.20866580307483673,
+      "learning_rate": 3.7454278567980406e-06,
+      "loss": 0.6834501028060913,
+      "step": 650
+    },
+    {
+      "epoch": 2.597609561752988,
+      "grad_norm": 0.7095727324485779,
+      "learning_rate": 3.7181455538609577e-06,
+      "loss": 0.6853353977203369,
+      "step": 652
+    },
+    {
+      "epoch": 2.605577689243028,
+      "grad_norm": 0.27560555934906006,
+      "learning_rate": 3.690940705055163e-06,
+      "loss": 0.6427911520004272,
+      "step": 654
+    },
+    {
+      "epoch": 2.6135458167330676,
+      "grad_norm": 0.2898000180721283,
+      "learning_rate": 3.6638144929283824e-06,
+      "loss": 0.29773062467575073,
+      "step": 656
+    },
+    {
+      "epoch": 2.6215139442231075,
+      "grad_norm": 0.33115100860595703,
+      "learning_rate": 3.6367680966101483e-06,
+      "loss": 0.5275067090988159,
+      "step": 658
+    },
+    {
+      "epoch": 2.6294820717131473,
+      "grad_norm": 0.337249755859375,
+      "learning_rate": 3.6098026917605334e-06,
+      "loss": 0.7844617962837219,
+      "step": 660
+    },
+    {
+      "epoch": 2.637450199203187,
+      "grad_norm": 1.0031529664993286,
+      "learning_rate": 3.5829194505190534e-06,
+      "loss": 0.7777225971221924,
+      "step": 662
+    },
+    {
+      "epoch": 2.645418326693227,
+      "grad_norm": 0.5599858164787292,
+      "learning_rate": 3.5561195414537164e-06,
+      "loss": 0.5193866491317749,
+      "step": 664
+    },
+    {
+      "epoch": 2.653386454183267,
+      "grad_norm": 0.46595117449760437,
+      "learning_rate": 3.529404129510221e-06,
+      "loss": 0.8348689079284668,
+      "step": 666
+    },
+    {
+      "epoch": 2.6613545816733066,
+      "grad_norm": 0.1848476678133011,
+      "learning_rate": 3.502774375961321e-06,
+      "loss": 0.5854208469390869,
+      "step": 668
+    },
+    {
+      "epoch": 2.6693227091633465,
+      "grad_norm": 0.8119276165962219,
+      "learning_rate": 3.4762314383563533e-06,
+      "loss": 0.627388060092926,
+      "step": 670
+    },
+    {
+      "epoch": 2.6772908366533863,
+      "grad_norm": 0.10560823231935501,
+      "learning_rate": 3.4497764704709143e-06,
+      "loss": 0.3013390302658081,
+      "step": 672
+    },
+    {
+      "epoch": 2.685258964143426,
+      "grad_norm": 0.42232856154441833,
+      "learning_rate": 3.4234106222567055e-06,
+      "loss": 0.9308726787567139,
+      "step": 674
+    },
+    {
+      "epoch": 2.6932270916334664,
+      "grad_norm": 1.0065451860427856,
+      "learning_rate": 3.3971350397915504e-06,
+      "loss": 0.6211099624633789,
+      "step": 676
+    },
+    {
+      "epoch": 2.7011952191235062,
+      "grad_norm": 0.5451958179473877,
+      "learning_rate": 3.370950865229583e-06,
+      "loss": 0.5485705137252808,
+      "step": 678
+    },
+    {
+      "epoch": 2.709163346613546,
+      "grad_norm": 0.35996702313423157,
+      "learning_rate": 3.344859236751585e-06,
+      "loss": 0.9292018413543701,
+      "step": 680
+    },
+    {
+      "epoch": 2.717131474103586,
+      "grad_norm": 0.27155205607414246,
+      "learning_rate": 3.3188612885155257e-06,
+      "loss": 0.9513242840766907,
+      "step": 682
+    },
+    {
+      "epoch": 2.7250996015936257,
+      "grad_norm": 0.5599476099014282,
+      "learning_rate": 3.2929581506072596e-06,
+      "loss": 0.6545392870903015,
+      "step": 684
+    },
+    {
+      "epoch": 2.7330677290836656,
+      "grad_norm": 0.5605886578559875,
+      "learning_rate": 3.267150948991387e-06,
+      "loss": 0.4231800138950348,
+      "step": 686
+    },
+    {
+      "epoch": 2.7410358565737054,
+      "grad_norm": 0.3724638521671295,
+      "learning_rate": 3.241440805462337e-06,
+      "loss": 0.9280582070350647,
+      "step": 688
+    },
+    {
+      "epoch": 2.7490039840637452,
+      "grad_norm": 0.4514469802379608,
+      "learning_rate": 3.215828837595585e-06,
+      "loss": 0.47799623012542725,
+      "step": 690
+    },
+    {
+      "epoch": 2.756972111553785,
+      "grad_norm": 0.33316367864608765,
+      "learning_rate": 3.190316158699084e-06,
+      "loss": 1.0304505825042725,
+      "step": 692
+    },
+    {
+      "epoch": 2.764940239043825,
+      "grad_norm": 0.6290820837020874,
+      "learning_rate": 3.1649038777648622e-06,
+      "loss": 0.7659841775894165,
+      "step": 694
+    },
+    {
+      "epoch": 2.7729083665338647,
+      "grad_norm": 0.6772921681404114,
+      "learning_rate": 3.1395930994208234e-06,
+      "loss": 0.5363284945487976,
+      "step": 696
+    },
+    {
+      "epoch": 2.7808764940239046,
+      "grad_norm": 0.24997524917125702,
+      "learning_rate": 3.114384923882734e-06,
+      "loss": 0.7607216835021973,
+      "step": 698
+    },
+    {
+      "epoch": 2.7888446215139444,
+      "grad_norm": 0.4008147716522217,
+      "learning_rate": 3.089280446906387e-06,
+      "loss": 0.9632657766342163,
+      "step": 700
+    },
+    {
+      "epoch": 2.7968127490039842,
+      "grad_norm": 0.2486056089401245,
+      "learning_rate": 3.0642807597399847e-06,
+      "loss": 0.5509517192840576,
+      "step": 702
+    },
+    {
+      "epoch": 2.804780876494024,
+      "grad_norm": 0.77519291639328,
+      "learning_rate": 3.0393869490766913e-06,
+      "loss": 1.048142910003662,
+      "step": 704
+    },
+    {
+      "epoch": 2.812749003984064,
+      "grad_norm": 0.35780569911003113,
+      "learning_rate": 3.0146000970074087e-06,
+      "loss": 1.0227577686309814,
+      "step": 706
+    },
+    {
+      "epoch": 2.8207171314741037,
+      "grad_norm": 0.4452541768550873,
+      "learning_rate": 2.9899212809737275e-06,
+      "loss": 1.0011272430419922,
+      "step": 708
+    },
+    {
+      "epoch": 2.8286852589641436,
+      "grad_norm": 0.29877766966819763,
+      "learning_rate": 2.9653515737211037e-06,
+      "loss": 0.4837097227573395,
+      "step": 710
+    },
+    {
+      "epoch": 2.8366533864541834,
+      "grad_norm": 0.2995551824569702,
+      "learning_rate": 2.940892043252225e-06,
+      "loss": 0.9765815138816833,
+      "step": 712
+    },
+    {
+      "epoch": 2.8446215139442232,
+      "grad_norm": 0.2545861303806305,
+      "learning_rate": 2.916543752780574e-06,
+      "loss": 0.5666689872741699,
+      "step": 714
+    },
+    {
+      "epoch": 2.852589641434263,
+      "grad_norm": 0.26393797993659973,
+      "learning_rate": 2.892307760684234e-06,
+      "loss": 0.9279859066009521,
+      "step": 716
+    },
+    {
+      "epoch": 2.860557768924303,
+      "grad_norm": 1.1712604761123657,
+      "learning_rate": 2.8681851204598675e-06,
+      "loss": 0.5682026743888855,
+      "step": 718
+    },
+    {
+      "epoch": 2.8685258964143427,
+      "grad_norm": 0.20210614800453186,
+      "learning_rate": 2.844176880676931e-06,
+      "loss": 0.5432709455490112,
+      "step": 720
+    },
+    {
+      "epoch": 2.8764940239043826,
+      "grad_norm": 0.4568478763103485,
+      "learning_rate": 2.8202840849320867e-06,
+      "loss": 0.6219796538352966,
+      "step": 722
+    },
+    {
+      "epoch": 2.8844621513944224,
+      "grad_norm": 0.24569456279277802,
+      "learning_rate": 2.7965077718038437e-06,
+      "loss": 0.6062704920768738,
+      "step": 724
+    },
+    {
+      "epoch": 2.8924302788844622,
+      "grad_norm": 0.42229318618774414,
+      "learning_rate": 2.7728489748074193e-06,
+      "loss": 0.8858544826507568,
+      "step": 726
+    },
+    {
+      "epoch": 2.900398406374502,
+      "grad_norm": 0.3378986120223999,
+      "learning_rate": 2.7493087223498015e-06,
+      "loss": 0.5284279584884644,
+      "step": 728
+    },
+    {
+      "epoch": 2.908366533864542,
+      "grad_norm": 0.5533449649810791,
+      "learning_rate": 2.725888037685057e-06,
+      "loss": 0.9063352346420288,
+      "step": 730
+    },
+    {
+      "epoch": 2.9163346613545817,
+      "grad_norm": 0.24462848901748657,
+      "learning_rate": 2.702587938869842e-06,
+      "loss": 0.2906056046485901,
+      "step": 732
+    },
+    {
+      "epoch": 2.9243027888446216,
+      "grad_norm": 0.26387494802474976,
+      "learning_rate": 2.6794094387191617e-06,
+      "loss": 0.9532904028892517,
+      "step": 734
+    },
+    {
+      "epoch": 2.9322709163346614,
+      "grad_norm": 0.8418635725975037,
+      "learning_rate": 2.6563535447623302e-06,
+      "loss": 0.40890583395957947,
+      "step": 736
+    },
+    {
+      "epoch": 2.9402390438247012,
+      "grad_norm": 0.9646982550621033,
+      "learning_rate": 2.6334212591991847e-06,
+      "loss": 0.7772600650787354,
+      "step": 738
+    },
+    {
+      "epoch": 2.948207171314741,
+      "grad_norm": 0.34290510416030884,
+      "learning_rate": 2.6106135788565248e-06,
+      "loss": 0.9934842586517334,
+      "step": 740
+    },
+    {
+      "epoch": 2.956175298804781,
+      "grad_norm": 0.8493300676345825,
+      "learning_rate": 2.5879314951447688e-06,
+      "loss": 0.25851064920425415,
+      "step": 742
+    },
+    {
+      "epoch": 2.9641434262948207,
+      "grad_norm": 0.32900604605674744,
+      "learning_rate": 2.5653759940148692e-06,
+      "loss": 0.7729766368865967,
+      "step": 744
+    },
+    {
+      "epoch": 2.9721115537848606,
+      "grad_norm": 0.5969822406768799,
+      "learning_rate": 2.542948055915454e-06,
+      "loss": 0.8053131103515625,
+      "step": 746
+    },
+    {
+      "epoch": 2.9800796812749004,
+      "grad_norm": 0.5320551991462708,
+      "learning_rate": 2.5206486557502087e-06,
+      "loss": 0.45959070324897766,
+      "step": 748
+    },
+    {
+      "epoch": 2.9880478087649402,
+      "grad_norm": 0.43073588609695435,
+      "learning_rate": 2.498478762835492e-06,
+      "loss": 0.12302790582180023,
+      "step": 750
+    },
+    {
+      "epoch": 2.99601593625498,
+      "grad_norm": 0.218783900141716,
+      "learning_rate": 2.476439340858208e-06,
+      "loss": 0.9228564500808716,
+      "step": 752
+    },
+    {
+      "epoch": 3.00398406374502,
+      "grad_norm": 0.3783073127269745,
+      "learning_rate": 2.4545313478339178e-06,
+      "loss": 0.4805603325366974,
+      "step": 754
+    },
+    {
+      "epoch": 3.0119521912350598,
+      "grad_norm": 0.32770872116088867,
+      "learning_rate": 2.4327557360651895e-06,
+      "loss": 0.5052729845046997,
+      "step": 756
+    },
+    {
+      "epoch": 3.0199203187250996,
+      "grad_norm": 0.7116933465003967,
+      "learning_rate": 2.411113452100211e-06,
+      "loss": 0.2500203847885132,
+      "step": 758
+    },
+    {
+      "epoch": 3.0278884462151394,
+      "grad_norm": 0.490122526884079,
+      "learning_rate": 2.3896054366916365e-06,
+      "loss": 0.5222545266151428,
+      "step": 760
+    },
+    {
+      "epoch": 3.0358565737051793,
+      "grad_norm": 0.37244126200675964,
+      "learning_rate": 2.3682326247557e-06,
+      "loss": 0.36726564168930054,
+      "step": 762
+    },
+    {
+      "epoch": 3.043824701195219,
+      "grad_norm": 0.2626785337924957,
+      "learning_rate": 2.346995945331576e-06,
+      "loss": 0.7889100909233093,
+      "step": 764
+    },
+    {
+      "epoch": 3.051792828685259,
+      "grad_norm": 0.2396881878376007,
+      "learning_rate": 2.325896321540993e-06,
+      "loss": 0.15941597521305084,
+      "step": 766
+    },
+    {
+      "epoch": 3.0597609561752988,
+      "grad_norm": 0.3975837230682373,
+      "learning_rate": 2.3049346705481086e-06,
+      "loss": 0.4646088778972626,
+      "step": 768
+    },
+    {
+      "epoch": 3.0677290836653386,
+      "grad_norm": 0.4999510943889618,
+      "learning_rate": 2.284111903519641e-06,
+      "loss": 0.6154494881629944,
+      "step": 770
+    },
+    {
+      "epoch": 3.0756972111553784,
+      "grad_norm": 0.385384202003479,
+      "learning_rate": 2.26342892558526e-06,
+      "loss": 0.7881206274032593,
+      "step": 772
+    },
+    {
+      "epoch": 3.0836653386454183,
+      "grad_norm": 0.6941497325897217,
+      "learning_rate": 2.2428866357982515e-06,
+      "loss": 0.7140223383903503,
+      "step": 774
+    },
+    {
+      "epoch": 3.091633466135458,
+      "grad_norm": 0.04772397503256798,
+      "learning_rate": 2.222485927096424e-06,
+      "loss": 0.43760406970977783,
+      "step": 776
+    },
+    {
+      "epoch": 3.099601593625498,
+      "grad_norm": 0.8838547468185425,
+      "learning_rate": 2.2022276862633075e-06,
+      "loss": 0.5001354813575745,
+      "step": 778
+    },
+    {
+      "epoch": 3.1075697211155378,
+      "grad_norm": 0.6882412433624268,
+      "learning_rate": 2.1821127938895938e-06,
+      "loss": 0.7709897756576538,
+      "step": 780
+    },
+    {
+      "epoch": 3.1155378486055776,
+      "grad_norm": 0.10992899537086487,
+      "learning_rate": 2.162142124334871e-06,
+      "loss": 0.1869785487651825,
+      "step": 782
+    },
+    {
+      "epoch": 3.1235059760956174,
+      "grad_norm": 0.23556940257549286,
+      "learning_rate": 2.1423165456896068e-06,
+      "loss": 0.8461598753929138,
+      "step": 784
+    },
+    {
+      "epoch": 3.1314741035856573,
+      "grad_norm": 0.27568379044532776,
+      "learning_rate": 2.1226369197374176e-06,
+      "loss": 0.6046760678291321,
+      "step": 786
+    },
+    {
+      "epoch": 3.139442231075697,
+      "grad_norm": 0.7430327534675598,
+      "learning_rate": 2.103104101917615e-06,
+      "loss": 0.4158598780632019,
+      "step": 788
+    },
+    {
+      "epoch": 3.147410358565737,
+      "grad_norm": 0.48514223098754883,
+      "learning_rate": 2.0837189412880044e-06,
+      "loss": 0.40696510672569275,
+      "step": 790
+    },
+    {
+      "epoch": 3.1553784860557768,
+      "grad_norm": 0.8990712761878967,
+      "learning_rate": 2.0644822804879964e-06,
+      "loss": 0.5259879231452942,
+      "step": 792
+    },
+    {
+      "epoch": 3.1633466135458166,
+      "grad_norm": 0.08627250045537949,
+      "learning_rate": 2.045394955701969e-06,
+      "loss": 0.16185329854488373,
+      "step": 794
+    },
+    {
+      "epoch": 3.1713147410358564,
+      "grad_norm": 0.6621522307395935,
+      "learning_rate": 2.0264577966229222e-06,
+      "loss": 0.4720742106437683,
+      "step": 796
+    },
+    {
+      "epoch": 3.1792828685258963,
+      "grad_norm": 0.4478707015514374,
+      "learning_rate": 2.007671626416412e-06,
+      "loss": 0.4690909683704376,
+      "step": 798
+    },
+    {
+      "epoch": 3.187250996015936,
+      "grad_norm": 1.3150907754898071,
+      "learning_rate": 1.9890372616847663e-06,
+      "loss": 0.6778526306152344,
+      "step": 800
+    },
+    {
+      "epoch": 3.195219123505976,
+      "grad_norm": 0.29304519295692444,
+      "learning_rate": 1.9705555124315965e-06,
+      "loss": 0.5810800194740295,
+      "step": 802
+    },
+    {
+      "epoch": 3.2031872509960158,
+      "grad_norm": 0.713803231716156,
+      "learning_rate": 1.95222718202658e-06,
+      "loss": 0.44393956661224365,
+      "step": 804
+    },
+    {
+      "epoch": 3.2111553784860556,
+      "grad_norm": 1.095471739768982,
+      "learning_rate": 1.934053067170545e-06,
+      "loss": 0.822306752204895,
+      "step": 806
+    },
+    {
+      "epoch": 3.2191235059760954,
+      "grad_norm": 0.714267909526825,
+      "learning_rate": 1.9160339578608305e-06,
+      "loss": 0.18660493195056915,
+      "step": 808
+    },
+    {
+      "epoch": 3.2270916334661353,
+      "grad_norm": 0.322010338306427,
+      "learning_rate": 1.8981706373569602e-06,
+      "loss": 0.550746738910675,
+      "step": 810
+    },
+    {
+      "epoch": 3.235059760956175,
+      "grad_norm": 1.6815509796142578,
+      "learning_rate": 1.8804638821465821e-06,
+      "loss": 0.32248857617378235,
+      "step": 812
+    },
+    {
+      "epoch": 3.243027888446215,
+      "grad_norm": 0.6803773045539856,
+      "learning_rate": 1.8629144619117234e-06,
+      "loss": 0.6694077253341675,
+      "step": 814
+    },
+    {
+      "epoch": 3.2509960159362548,
+      "grad_norm": 1.0847859382629395,
+      "learning_rate": 1.8455231394953348e-06,
+      "loss": 0.20106957852840424,
+      "step": 816
+    },
+    {
+      "epoch": 3.2589641434262946,
+      "grad_norm": 0.2348478138446808,
+      "learning_rate": 1.8282906708681238e-06,
+      "loss": 0.7628844976425171,
+      "step": 818
+    },
+    {
+      "epoch": 3.2669322709163344,
+      "grad_norm": 5.2003560066223145,
+      "learning_rate": 1.811217805095699e-06,
+      "loss": 0.35256174206733704,
+      "step": 820
+    },
+    {
+      "epoch": 3.2749003984063743,
+      "grad_norm": 0.856585681438446,
+      "learning_rate": 1.7943052843060133e-06,
+      "loss": 0.4454050660133362,
+      "step": 822
+    },
+    {
+      "epoch": 3.2828685258964145,
+      "grad_norm": 10.908398628234863,
+      "learning_rate": 1.7775538436570937e-06,
+      "loss": 0.44718843698501587,
+      "step": 824
+    },
+    {
+      "epoch": 3.2908366533864544,
+      "grad_norm": 0.2760160565376282,
+      "learning_rate": 1.7609642113050996e-06,
+      "loss": 0.771109938621521,
+      "step": 826
+    },
+    {
+      "epoch": 3.298804780876494,
+      "grad_norm": 0.30608290433883667,
+      "learning_rate": 1.7445371083726567e-06,
+      "loss": 0.7913384437561035,
+      "step": 828
+    },
+    {
+      "epoch": 3.306772908366534,
+      "grad_norm": 0.5408643484115601,
+      "learning_rate": 1.7282732489175224e-06,
+      "loss": 0.6727736592292786,
+      "step": 830
+    },
+    {
+      "epoch": 3.314741035856574,
+      "grad_norm": 0.8327075839042664,
+      "learning_rate": 1.7121733399015406e-06,
+      "loss": 0.36132076382637024,
+      "step": 832
+    },
+    {
+      "epoch": 3.3227091633466137,
+      "grad_norm": 1.3813250064849854,
+      "learning_rate": 1.6962380811599138e-06,
+      "loss": 0.23368409276008606,
+      "step": 834
+    },
+    {
+      "epoch": 3.3306772908366535,
+      "grad_norm": 1.03911292552948,
+      "learning_rate": 1.6804681653707816e-06,
+      "loss": 0.4291265904903412,
+      "step": 836
+    },
+    {
+      "epoch": 3.3386454183266934,
+      "grad_norm": 0.3465489447116852,
+      "learning_rate": 1.6648642780251096e-06,
+      "loss": 0.594310462474823,
+      "step": 838
+    },
+    {
+      "epoch": 3.346613545816733,
+      "grad_norm": 0.7986364960670471,
+      "learning_rate": 1.6494270973968974e-06,
+      "loss": 0.1275523155927658,
+      "step": 840
+    },
+    {
+      "epoch": 3.354581673306773,
+      "grad_norm": 0.4420923590660095,
+      "learning_rate": 1.6341572945136902e-06,
+      "loss": 0.4623723030090332,
+      "step": 842
+    },
+    {
+      "epoch": 3.362549800796813,
+      "grad_norm": 0.34837982058525085,
+      "learning_rate": 1.6190555331274122e-06,
+      "loss": 0.47705015540122986,
+      "step": 844
+    },
+    {
+      "epoch": 3.3705179282868527,
+      "grad_norm": 0.32239773869514465,
+      "learning_rate": 1.6041224696855145e-06,
+      "loss": 0.8339754939079285,
+      "step": 846
+    },
+    {
+      "epoch": 3.3784860557768925,
+      "grad_norm": 4.646689414978027,
+      "learning_rate": 1.589358753302439e-06,
+      "loss": 0.7510449886322021,
+      "step": 848
+    },
+    {
+      "epoch": 3.3864541832669324,
+      "grad_norm": 0.8043104410171509,
+      "learning_rate": 1.5747650257314068e-06,
+      "loss": 0.6570635437965393,
+      "step": 850
+    },
+    {
+      "epoch": 3.394422310756972,
+      "grad_norm": 0.28592732548713684,
+      "learning_rate": 1.5603419213365146e-06,
+      "loss": 0.7621712684631348,
+      "step": 852
+    },
+    {
+      "epoch": 3.402390438247012,
+      "grad_norm": 0.6587794423103333,
+      "learning_rate": 1.5460900670651705e-06,
+      "loss": 0.4589743912220001,
+      "step": 854
+    },
+    {
+      "epoch": 3.410358565737052,
+      "grad_norm": 0.3793962597846985,
+      "learning_rate": 1.5320100824208306e-06,
+      "loss": 0.7107986211776733,
+      "step": 856
+    },
+    {
+      "epoch": 3.4183266932270917,
+      "grad_norm": 0.3333897888660431,
+      "learning_rate": 1.51810257943608e-06,
+      "loss": 0.7919248938560486,
+      "step": 858
+    },
+    {
+      "epoch": 3.4262948207171315,
+      "grad_norm": 0.4181341528892517,
+      "learning_rate": 1.5043681626460196e-06,
+      "loss": 0.8178535103797913,
+      "step": 860
+    },
+    {
+      "epoch": 3.4342629482071714,
+      "grad_norm": 0.25965699553489685,
+      "learning_rate": 1.4908074290619972e-06,
+      "loss": 0.5096842050552368,
+      "step": 862
+    },
+    {
+      "epoch": 3.442231075697211,
+      "grad_norm": 0.6406610012054443,
+      "learning_rate": 1.477420968145652e-06,
+      "loss": 0.3832019567489624,
+      "step": 864
+    },
+    {
+      "epoch": 3.450199203187251,
+      "grad_norm": 0.26335206627845764,
+      "learning_rate": 1.4642093617832853e-06,
+      "loss": 0.6468888521194458,
+      "step": 866
+    },
+    {
+      "epoch": 3.458167330677291,
+      "grad_norm": 0.34508153796195984,
+      "learning_rate": 1.4511731842605805e-06,
+      "loss": 0.6055460572242737,
+      "step": 868
+    },
+    {
+      "epoch": 3.4661354581673307,
+      "grad_norm": 0.44761452078819275,
+      "learning_rate": 1.4383130022376298e-06,
+      "loss": 0.44755107164382935,
+      "step": 870
+    },
+    {
+      "epoch": 3.4741035856573705,
+      "grad_norm": 0.11638348549604416,
+      "learning_rate": 1.4256293747243065e-06,
+      "loss": 0.4698525369167328,
+      "step": 872
+    },
+    {
+      "epoch": 3.4820717131474104,
+      "grad_norm": 0.13195247948169708,
+      "learning_rate": 1.4131228530559637e-06,
+      "loss": 0.5005823373794556,
+      "step": 874
+    },
+    {
+      "epoch": 3.49003984063745,
+      "grad_norm": 1.5164440870285034,
+      "learning_rate": 1.400793980869468e-06,
+      "loss": 0.40101391077041626,
+      "step": 876
+    },
+    {
+      "epoch": 3.49800796812749,
+      "grad_norm": 0.2557206153869629,
+      "learning_rate": 1.3886432940795744e-06,
+      "loss": 0.4368343651294708,
+      "step": 878
+    },
+    {
+      "epoch": 3.50597609561753,
+      "grad_norm": 0.15385140478610992,
+      "learning_rate": 1.3766713208556232e-06,
+      "loss": 0.27624547481536865,
+      "step": 880
+    },
+    {
+      "epoch": 3.5139442231075697,
+      "grad_norm": 0.613798975944519,
+      "learning_rate": 1.3648785815985876e-06,
+      "loss": 0.36860376596450806,
+      "step": 882
+    },
+    {
+      "epoch": 3.5219123505976095,
+      "grad_norm": 0.5025630593299866,
+      "learning_rate": 1.353265588918448e-06,
+      "loss": 0.06519927084445953,
+      "step": 884
+    },
+    {
+      "epoch": 3.5298804780876494,
+      "grad_norm": 1.1982804536819458,
+      "learning_rate": 1.3418328476119136e-06,
+      "loss": 0.6289451718330383,
+      "step": 886
+    },
+    {
+      "epoch": 3.537848605577689,
+      "grad_norm": 0.6196771264076233,
+      "learning_rate": 1.330580854640477e-06,
+      "loss": 0.7229180335998535,
+      "step": 888
+    },
+    {
+      "epoch": 3.545816733067729,
+      "grad_norm": 0.3435492515563965,
+      "learning_rate": 1.3195100991088142e-06,
+      "loss": 0.41303932666778564,
+      "step": 890
+    },
+    {
+      "epoch": 3.553784860557769,
+      "grad_norm": 1.2850234508514404,
+      "learning_rate": 1.308621062243524e-06,
+      "loss": 0.6361463069915771,
+      "step": 892
+    },
+    {
+      "epoch": 3.5617529880478087,
+      "grad_norm": 0.6895233988761902,
+      "learning_rate": 1.2979142173722054e-06,
+      "loss": 0.9105860590934753,
+      "step": 894
+    },
+    {
+      "epoch": 3.5697211155378485,
+      "grad_norm": 0.38753747940063477,
+      "learning_rate": 1.2873900299028896e-06,
+      "loss": 0.15039730072021484,
+      "step": 896
+    },
+    {
+      "epoch": 3.5776892430278884,
+      "grad_norm": 0.2424442321062088,
+      "learning_rate": 1.2770489573038073e-06,
+      "loss": 0.4806039035320282,
+      "step": 898
+    },
+    {
+      "epoch": 3.585657370517928,
+      "grad_norm": 0.3263397216796875,
+      "learning_rate": 1.2668914490834983e-06,
+      "loss": 0.13606518507003784,
+      "step": 900
+    },
+    {
+      "epoch": 3.593625498007968,
+      "grad_norm": 1.051065444946289,
+      "learning_rate": 1.2569179467712805e-06,
+      "loss": 0.08013846725225449,
+      "step": 902
+    },
+    {
+      "epoch": 3.601593625498008,
+      "grad_norm": 0.45325273275375366,
+      "learning_rate": 1.2471288838980505e-06,
+      "loss": 0.5731567144393921,
+      "step": 904
+    },
+    {
+      "epoch": 3.6095617529880477,
+      "grad_norm": 0.49484413862228394,
+      "learning_rate": 1.237524685977444e-06,
+      "loss": 0.44954267144203186,
+      "step": 906
+    },
+    {
+      "epoch": 3.6175298804780875,
+      "grad_norm": 0.385636568069458,
+      "learning_rate": 1.2281057704873331e-06,
+      "loss": 0.701659619808197,
+      "step": 908
+    },
+    {
+      "epoch": 3.6254980079681274,
+      "grad_norm": 0.47307252883911133,
+      "learning_rate": 1.2188725468516876e-06,
+      "loss": 0.6697795987129211,
+      "step": 910
+    },
+    {
+      "epoch": 3.633466135458167,
+      "grad_norm": 0.08213522285223007,
+      "learning_rate": 1.2098254164227706e-06,
+      "loss": 0.024407412856817245,
+      "step": 912
+    },
+    {
+      "epoch": 3.641434262948207,
+      "grad_norm": 0.3121781051158905,
+      "learning_rate": 1.200964772463695e-06,
+      "loss": 0.5857076048851013,
+      "step": 914
+    },
+    {
+      "epoch": 3.649402390438247,
+      "grad_norm": 0.7764087915420532,
+      "learning_rate": 1.1922910001313312e-06,
+      "loss": 0.6128016710281372,
+      "step": 916
+    },
+    {
+      "epoch": 3.6573705179282867,
+      "grad_norm": 0.9828333258628845,
+      "learning_rate": 1.1838044764595617e-06,
+      "loss": 0.10074111074209213,
+      "step": 918
+    },
+    {
+      "epoch": 3.6653386454183265,
+      "grad_norm": 0.3853728175163269,
+      "learning_rate": 1.1755055703428958e-06,
+      "loss": 0.7550044655799866,
+      "step": 920
+    },
+    {
+      "epoch": 3.6733067729083664,
+      "grad_norm": 0.381106436252594,
+      "learning_rate": 1.1673946425204288e-06,
+      "loss": 0.09400598704814911,
+      "step": 922
+    },
+    {
+      "epoch": 3.681274900398406,
+      "grad_norm": 0.9098880290985107,
+      "learning_rate": 1.1594720455601664e-06,
+      "loss": 0.7232218980789185,
+      "step": 924
+    },
+    {
+      "epoch": 3.6892430278884465,
+      "grad_norm": 0.7898821830749512,
+      "learning_rate": 1.1517381238436986e-06,
+      "loss": 0.1783030927181244,
+      "step": 926
+    },
+    {
+      "epoch": 3.6972111553784863,
+      "grad_norm": 0.3300727605819702,
+      "learning_rate": 1.1441932135512262e-06,
+      "loss": 0.42349258065223694,
+      "step": 928
+    },
+    {
+      "epoch": 3.705179282868526,
+      "grad_norm": 0.15504029393196106,
+      "learning_rate": 1.1368376426469526e-06,
+      "loss": 0.6427653431892395,
+      "step": 930
+    },
+    {
+      "epoch": 3.713147410358566,
+      "grad_norm": 0.33219656348228455,
+      "learning_rate": 1.1296717308648237e-06,
+      "loss": 0.7705016732215881,
+      "step": 932
+    },
+    {
+      "epoch": 3.721115537848606,
+      "grad_norm": 0.4385871887207031,
+      "learning_rate": 1.1226957896946332e-06,
+      "loss": 0.3702158033847809,
+      "step": 934
+    },
+    {
+      "epoch": 3.7290836653386457,
+      "grad_norm": 0.3322261869907379,
+      "learning_rate": 1.115910122368479e-06,
+      "loss": 0.6915496587753296,
+      "step": 936
+    },
+    {
+      "epoch": 3.7370517928286855,
+      "grad_norm": 2.8838603496551514,
+      "learning_rate": 1.1093150238475847e-06,
+      "loss": 0.35162052512168884,
+      "step": 938
+    },
+    {
+      "epoch": 3.7450199203187253,
+      "grad_norm": 0.408324658870697,
+      "learning_rate": 1.1029107808094789e-06,
+      "loss": 0.6292784810066223,
+      "step": 940
+    },
+    {
+      "epoch": 3.752988047808765,
+      "grad_norm": 1.1569932699203491,
+      "learning_rate": 1.0966976716355294e-06,
+      "loss": 0.4764218330383301,
+      "step": 942
+    },
+    {
+      "epoch": 3.760956175298805,
+      "grad_norm": 0.20497426390647888,
+      "learning_rate": 1.0906759663988475e-06,
+      "loss": 0.43546736240386963,
+      "step": 944
+    },
+    {
+      "epoch": 3.768924302788845,
+      "grad_norm": 0.316875159740448,
+      "learning_rate": 1.084845926852547e-06,
+      "loss": 0.8292940258979797,
+      "step": 946
+    },
+    {
+      "epoch": 3.7768924302788847,
+      "grad_norm": 0.5094802379608154,
+      "learning_rate": 1.0792078064183652e-06,
+      "loss": 0.4230027496814728,
+      "step": 948
+    },
+    {
+      "epoch": 3.7848605577689245,
+      "grad_norm": 0.2979755699634552,
+      "learning_rate": 1.0737618501756455e-06,
+      "loss": 0.7801587581634521,
+      "step": 950
+    },
+    {
+      "epoch": 3.7928286852589643,
+      "grad_norm": 0.34857067465782166,
+      "learning_rate": 1.0685082948506875e-06,
+      "loss": 0.7767475843429565,
+      "step": 952
+    },
+    {
+      "epoch": 3.800796812749004,
+      "grad_norm": 1.9035229682922363,
+      "learning_rate": 1.063447368806457e-06,
+      "loss": 0.4040529727935791,
+      "step": 954
+    },
+    {
+      "epoch": 3.808764940239044,
+      "grad_norm": 0.5326647162437439,
+      "learning_rate": 1.0585792920326565e-06,
+      "loss": 0.607650637626648,
+      "step": 956
+    },
+    {
+      "epoch": 3.816733067729084,
+      "grad_norm": 0.6186449527740479,
+      "learning_rate": 1.0539042761361646e-06,
+      "loss": 0.38428518176078796,
+      "step": 958
+    },
+    {
+      "epoch": 3.8247011952191237,
+      "grad_norm": 1.3463289737701416,
+      "learning_rate": 1.0494225243318373e-06,
+      "loss": 0.42564353346824646,
+      "step": 960
+    },
+    {
+      "epoch": 3.8326693227091635,
+      "grad_norm": 0.12071845680475235,
+      "learning_rate": 1.0451342314336755e-06,
+      "loss": 0.4637273848056793,
+      "step": 962
+    },
+    {
+      "epoch": 3.8406374501992033,
+      "grad_norm": 0.3585212230682373,
+      "learning_rate": 1.0410395838463547e-06,
+      "loss": 0.8156216144561768,
+      "step": 964
+    },
+    {
+      "epoch": 3.848605577689243,
+      "grad_norm": 0.9284526705741882,
+      "learning_rate": 1.0371387595571248e-06,
+      "loss": 0.251337468624115,
+      "step": 966
+    },
+    {
+      "epoch": 3.856573705179283,
+      "grad_norm": 0.34982913732528687,
+      "learning_rate": 1.0334319281280715e-06,
+      "loss": 0.491554319858551,
+      "step": 968
+    },
+    {
+      "epoch": 3.864541832669323,
+      "grad_norm": 0.4465697407722473,
+      "learning_rate": 1.0299192506887467e-06,
+      "loss": 0.5315884351730347,
+      "step": 970
+    },
+    {
+      "epoch": 3.8725099601593627,
+      "grad_norm": 0.12670713663101196,
+      "learning_rate": 1.0266008799291635e-06,
+      "loss": 0.4361288845539093,
+      "step": 972
+    },
+    {
+      "epoch": 3.8804780876494025,
+      "grad_norm": 0.1383129358291626,
+      "learning_rate": 1.0234769600931612e-06,
+      "loss": 0.35626745223999023,
+      "step": 974
+    },
+    {
+      "epoch": 3.8884462151394423,
+      "grad_norm": 0.2924402058124542,
+      "learning_rate": 1.0205476269721312e-06,
+      "loss": 0.7829856276512146,
+      "step": 976
+    },
+    {
+      "epoch": 3.896414342629482,
+      "grad_norm": 1.4608038663864136,
+      "learning_rate": 1.0178130078991196e-06,
+      "loss": 0.4326501488685608,
+      "step": 978
+    },
+    {
+      "epoch": 3.904382470119522,
+      "grad_norm": 0.16665737330913544,
+      "learning_rate": 1.0152732217432875e-06,
+      "loss": 0.32169440388679504,
+      "step": 980
+    },
+    {
+      "epoch": 3.912350597609562,
+      "grad_norm": 0.6928649544715881,
+      "learning_rate": 1.0129283789047473e-06,
+      "loss": 0.30814969539642334,
+      "step": 982
+    },
+    {
+      "epoch": 3.9203187250996017,
+      "grad_norm": 0.26950931549072266,
+      "learning_rate": 1.0107785813097621e-06,
+      "loss": 0.732113778591156,
+      "step": 984
+    },
+    {
+      "epoch": 3.9282868525896415,
+      "grad_norm": 0.474765419960022,
+      "learning_rate": 1.0088239224063153e-06,
+      "loss": 0.15787996351718903,
+      "step": 986
+    },
+    {
+      "epoch": 3.9362549800796813,
+      "grad_norm": 0.9765795469284058,
+      "learning_rate": 1.0070644871600495e-06,
+      "loss": 0.5307015776634216,
+      "step": 988
+    },
+    {
+      "epoch": 3.944223107569721,
+      "grad_norm": 0.09540683031082153,
+      "learning_rate": 1.0055003520505714e-06,
+      "loss": 0.43287578225135803,
+      "step": 990
+    },
+    {
+      "epoch": 3.952191235059761,
+      "grad_norm": 0.39477550983428955,
+      "learning_rate": 1.00413158506813e-06,
+      "loss": 0.7727532982826233,
+      "step": 992
+    },
+    {
+      "epoch": 3.960159362549801,
+      "grad_norm": 0.17617295682430267,
+      "learning_rate": 1.0029582457106588e-06,
+      "loss": 0.18191753327846527,
+      "step": 994
+    },
+    {
+      "epoch": 3.9681274900398407,
+      "grad_norm": 1.1495691537857056,
+      "learning_rate": 1.00198038498119e-06,
+      "loss": 0.6728793382644653,
+      "step": 996
+    },
+    {
+      "epoch": 3.9760956175298805,
+      "grad_norm": 0.19389505684375763,
+      "learning_rate": 1.0011980453856391e-06,
+      "loss": 0.4585638642311096,
+      "step": 998
+    },
+    {
+      "epoch": 3.9840637450199203,
+      "grad_norm": 0.44044798612594604,
+      "learning_rate": 1.0006112609309548e-06,
+      "loss": 0.33413246273994446,
+      "step": 1000
+    },
+    {
+      "epoch": 3.99203187250996,
+      "grad_norm": 0.2927405536174774,
+      "learning_rate": 1.0002200571236429e-06,
+      "loss": 0.7803468704223633,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.281700074672699,
+      "learning_rate": 1.0000244509686564e-06,
+      "loss": 0.19255180656909943,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0,
+      "step": 1004,
+      "total_flos": 3.519329211045118e+18,
+      "train_loss": 0.8036329971025308,
+      "train_runtime": 8899.7642,
+      "train_samples_per_second": 6.769,
+      "train_steps_per_second": 0.113
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1004,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 99999,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.519329211045118e+18,
+  "train_batch_size": 3,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2dae75b55731660e84e9bdaf07e0f5add551266c5c06e6965e9cc1e4dcced9a
+size 5649