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 +1,63 @@
+---
+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_adamw_lr = {
+        "language": _fallback(getattr(training_args, "language_adamw_lr", 2.5e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_adamw_lr", 4e-6), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_adamw_lr", 1.5e-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 +1,63 @@
+---
+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_adamw_lr = {
+        "language": _fallback(getattr(training_args, "language_adamw_lr", 2.5e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_adamw_lr", 4e-6), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_adamw_lr", 1.5e-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": 7409.9375163599425,
+    "epoch": 4.0,
+    "total_flos": 4.038502240003031e+18,
+    "train_loss": 0.7041097417681578,
+    "train_runtime": 10571.9353,
+    "train_samples_per_second": 5.698,
+    "train_steps_per_second": 0.095
+}

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 sha256:1b1d46fb0a341c32054ec02f14df4f6c837c0f737276c64b7bf5958c93c907da
+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": 7409.9375163599425,
+    "epoch": 4.0,
+    "total_flos": 4.038502240003031e+18,
+    "train_loss": 0.7041097417681578,
+    "train_runtime": 10571.9353,
+    "train_samples_per_second": 5.698,
+    "train_steps_per_second": 0.095
+}

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": 2.023987293243408,
+      "learning_rate": 4.901960784313725e-07,
+      "loss": 2.090776205062866,
+      "step": 2
+    },
+    {
+      "epoch": 0.01593625498007968,
+      "grad_norm": 0.7315244078636169,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 2.0096936225891113,
+      "step": 4
+    },
+    {
+      "epoch": 0.02390438247011952,
+      "grad_norm": 3.48097825050354,
+      "learning_rate": 2.450980392156863e-06,
+      "loss": 2.0481386184692383,
+      "step": 6
+    },
+    {
+      "epoch": 0.03187250996015936,
+      "grad_norm": 0.44064468145370483,
+      "learning_rate": 3.431372549019608e-06,
+      "loss": 1.7936124801635742,
+      "step": 8
+    },
+    {
+      "epoch": 0.0398406374501992,
+      "grad_norm": 0.7415902614593506,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 1.2419222593307495,
+      "step": 10
+    },
+    {
+      "epoch": 0.04780876494023904,
+      "grad_norm": 0.45377975702285767,
+      "learning_rate": 5.392156862745099e-06,
+      "loss": 1.536027193069458,
+      "step": 12
+    },
+    {
+      "epoch": 0.055776892430278883,
+      "grad_norm": 0.5584130883216858,
+      "learning_rate": 6.372549019607843e-06,
+      "loss": 1.552971601486206,
+      "step": 14
+    },
+    {
+      "epoch": 0.06374501992031872,
+      "grad_norm": 2.158409833908081,
+      "learning_rate": 7.3529411764705884e-06,
+      "loss": 1.408432126045227,
+      "step": 16
+    },
+    {
+      "epoch": 0.07171314741035857,
+      "grad_norm": 0.8985700607299805,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.5199083089828491,
+      "step": 18
+    },
+    {
+      "epoch": 0.0796812749003984,
+      "grad_norm": 0.6790184378623962,
+      "learning_rate": 9.31372549019608e-06,
+      "loss": 1.4455268383026123,
+      "step": 20
+    },
+    {
+      "epoch": 0.08764940239043825,
+      "grad_norm": 1.0950764417648315,
+      "learning_rate": 1.0294117647058824e-05,
+      "loss": 1.0664888620376587,
+      "step": 22
+    },
+    {
+      "epoch": 0.09561752988047809,
+      "grad_norm": 0.720383882522583,
+      "learning_rate": 1.1274509803921569e-05,
+      "loss": 1.2728652954101562,
+      "step": 24
+    },
+    {
+      "epoch": 0.10358565737051793,
+      "grad_norm": 0.5877739787101746,
+      "learning_rate": 1.2254901960784313e-05,
+      "loss": 1.3846259117126465,
+      "step": 26
+    },
+    {
+      "epoch": 0.11155378486055777,
+      "grad_norm": 0.6533491611480713,
+      "learning_rate": 1.323529411764706e-05,
+      "loss": 1.6152613162994385,
+      "step": 28
+    },
+    {
+      "epoch": 0.11952191235059761,
+      "grad_norm": 0.7580568194389343,
+      "learning_rate": 1.4215686274509804e-05,
+      "loss": 1.350507378578186,
+      "step": 30
+    },
+    {
+      "epoch": 0.12749003984063745,
+      "grad_norm": 0.9999676942825317,
+      "learning_rate": 1.5196078431372548e-05,
+      "loss": 1.1312873363494873,
+      "step": 32
+    },
+    {
+      "epoch": 0.13545816733067728,
+      "grad_norm": 1.4729697704315186,
+      "learning_rate": 1.6176470588235296e-05,
+      "loss": 1.2183302640914917,
+      "step": 34
+    },
+    {
+      "epoch": 0.14342629482071714,
+      "grad_norm": 0.3444910943508148,
+      "learning_rate": 1.715686274509804e-05,
+      "loss": 1.3639639616012573,
+      "step": 36
+    },
+    {
+      "epoch": 0.15139442231075698,
+      "grad_norm": 0.67720627784729,
+      "learning_rate": 1.8137254901960785e-05,
+      "loss": 1.1725502014160156,
+      "step": 38
+    },
+    {
+      "epoch": 0.1593625498007968,
+      "grad_norm": 0.46388429403305054,
+      "learning_rate": 1.9117647058823528e-05,
+      "loss": 1.2786129713058472,
+      "step": 40
+    },
+    {
+      "epoch": 0.16733067729083664,
+      "grad_norm": 1.526458978652954,
+      "learning_rate": 2.0098039215686277e-05,
+      "loss": 0.8266106843948364,
+      "step": 42
+    },
+    {
+      "epoch": 0.1752988047808765,
+      "grad_norm": 0.4335272014141083,
+      "learning_rate": 2.107843137254902e-05,
+      "loss": 0.9850126504898071,
+      "step": 44
+    },
+    {
+      "epoch": 0.18326693227091634,
+      "grad_norm": 1.5010402202606201,
+      "learning_rate": 2.2058823529411766e-05,
+      "loss": 0.9776768088340759,
+      "step": 46
+    },
+    {
+      "epoch": 0.19123505976095617,
+      "grad_norm": 0.4024801254272461,
+      "learning_rate": 2.303921568627451e-05,
+      "loss": 1.201263189315796,
+      "step": 48
+    },
+    {
+      "epoch": 0.199203187250996,
+      "grad_norm": 1.3974441289901733,
+      "learning_rate": 2.401960784313726e-05,
+      "loss": 1.3466823101043701,
+      "step": 50
+    },
+    {
+      "epoch": 0.20717131474103587,
+      "grad_norm": 0.4053344428539276,
+      "learning_rate": 2.5e-05,
+      "loss": 1.304817795753479,
+      "step": 52
+    },
+    {
+      "epoch": 0.2151394422310757,
+      "grad_norm": 1.178676724433899,
+      "learning_rate": 2.4999755490977716e-05,
+      "loss": 1.170669436454773,
+      "step": 54
+    },
+    {
+      "epoch": 0.22310756972111553,
+      "grad_norm": 0.8294330835342407,
+      "learning_rate": 2.4999021974539246e-05,
+      "loss": 1.5720685720443726,
+      "step": 56
+    },
+    {
+      "epoch": 0.23107569721115537,
+      "grad_norm": 0.4647572934627533,
+      "learning_rate": 2.499779948256928e-05,
+      "loss": 1.3848868608474731,
+      "step": 58
+    },
+    {
+      "epoch": 0.23904382470119523,
+      "grad_norm": 0.5197198987007141,
+      "learning_rate": 2.4996088068207436e-05,
+      "loss": 0.9066517353057861,
+      "step": 60
+    },
+    {
+      "epoch": 0.24701195219123506,
+      "grad_norm": 0.5595812797546387,
+      "learning_rate": 2.499388780584593e-05,
+      "loss": 1.3845858573913574,
+      "step": 62
+    },
+    {
+      "epoch": 0.2549800796812749,
+      "grad_norm": 1.0602760314941406,
+      "learning_rate": 2.499119879112636e-05,
+      "loss": 1.1874157190322876,
+      "step": 64
+    },
+    {
+      "epoch": 0.26294820717131473,
+      "grad_norm": 0.2792101204395294,
+      "learning_rate": 2.498802114093555e-05,
+      "loss": 1.3185009956359863,
+      "step": 66
+    },
+    {
+      "epoch": 0.27091633466135456,
+      "grad_norm": 2.6775331497192383,
+      "learning_rate": 2.498435499340046e-05,
+      "loss": 1.5893816947937012,
+      "step": 68
+    },
+    {
+      "epoch": 0.2788844621513944,
+      "grad_norm": 0.5143619775772095,
+      "learning_rate": 2.498020050788218e-05,
+      "loss": 1.6420989036560059,
+      "step": 70
+    },
+    {
+      "epoch": 0.2868525896414343,
+      "grad_norm": 0.6029499173164368,
+      "learning_rate": 2.4975557864969017e-05,
+      "loss": 0.9500645995140076,
+      "step": 72
+    },
+    {
+      "epoch": 0.2948207171314741,
+      "grad_norm": 0.6316478252410889,
+      "learning_rate": 2.497042726646862e-05,
+      "loss": 1.224064826965332,
+      "step": 74
+    },
+    {
+      "epoch": 0.30278884462151395,
+      "grad_norm": 0.3514266908168793,
+      "learning_rate": 2.496480893539924e-05,
+      "loss": 1.5448570251464844,
+      "step": 76
+    },
+    {
+      "epoch": 0.3107569721115538,
+      "grad_norm": 0.950763463973999,
+      "learning_rate": 2.4958703115980007e-05,
+      "loss": 1.0478010177612305,
+      "step": 78
+    },
+    {
+      "epoch": 0.3187250996015936,
+      "grad_norm": 0.2530899941921234,
+      "learning_rate": 2.4952110073620335e-05,
+      "loss": 0.9237808585166931,
+      "step": 80
+    },
+    {
+      "epoch": 0.32669322709163345,
+      "grad_norm": 0.32029739022254944,
+      "learning_rate": 2.4945030094908376e-05,
+      "loss": 1.4276543855667114,
+      "step": 82
+    },
+    {
+      "epoch": 0.3346613545816733,
+      "grad_norm": 0.5300575494766235,
+      "learning_rate": 2.493746348759856e-05,
+      "loss": 1.0075225830078125,
+      "step": 84
+    },
+    {
+      "epoch": 0.3426294820717131,
+      "grad_norm": 0.25728508830070496,
+      "learning_rate": 2.4929410580598218e-05,
+      "loss": 1.397177815437317,
+      "step": 86
+    },
+    {
+      "epoch": 0.350597609561753,
+      "grad_norm": 0.6185736060142517,
+      "learning_rate": 2.4920871723953288e-05,
+      "loss": 1.047032117843628,
+      "step": 88
+    },
+    {
+      "epoch": 0.35856573705179284,
+      "grad_norm": 0.3299442529678345,
+      "learning_rate": 2.4911847288833108e-05,
+      "loss": 1.3797515630722046,
+      "step": 90
+    },
+    {
+      "epoch": 0.3665338645418327,
+      "grad_norm": 1.0324299335479736,
+      "learning_rate": 2.4902337667514263e-05,
+      "loss": 1.2001453638076782,
+      "step": 92
+    },
+    {
+      "epoch": 0.3745019920318725,
+      "grad_norm": 0.26941925287246704,
+      "learning_rate": 2.4892343273363546e-05,
+      "loss": 1.2868791818618774,
+      "step": 94
+    },
+    {
+      "epoch": 0.38247011952191234,
+      "grad_norm": 1.6706314086914062,
+      "learning_rate": 2.4881864540819994e-05,
+      "loss": 1.144903302192688,
+      "step": 96
+    },
+    {
+      "epoch": 0.3904382470119522,
+      "grad_norm": 0.698556125164032,
+      "learning_rate": 2.487090192537598e-05,
+      "loss": 1.2919436693191528,
+      "step": 98
+    },
+    {
+      "epoch": 0.398406374501992,
+      "grad_norm": 0.25945818424224854,
+      "learning_rate": 2.485945590355744e-05,
+      "loss": 0.9402350187301636,
+      "step": 100
+    },
+    {
+      "epoch": 0.4063745019920319,
+      "grad_norm": 1.8866931200027466,
+      "learning_rate": 2.4847526972903148e-05,
+      "loss": 1.2891254425048828,
+      "step": 102
+    },
+    {
+      "epoch": 0.41434262948207173,
+      "grad_norm": 0.2679593563079834,
+      "learning_rate": 2.4835115651943096e-05,
+      "loss": 0.8440443873405457,
+      "step": 104
+    },
+    {
+      "epoch": 0.42231075697211157,
+      "grad_norm": 1.1126761436462402,
+      "learning_rate": 2.482222248017593e-05,
+      "loss": 0.7792216539382935,
+      "step": 106
+    },
+    {
+      "epoch": 0.4302788844621514,
+      "grad_norm": 0.40202105045318604,
+      "learning_rate": 2.4808848018045546e-05,
+      "loss": 1.046218752861023,
+      "step": 108
+    },
+    {
+      "epoch": 0.43824701195219123,
+      "grad_norm": 0.29720884561538696,
+      "learning_rate": 2.4794992846916674e-05,
+      "loss": 1.3036984205245972,
+      "step": 110
+    },
+    {
+      "epoch": 0.44621513944223107,
+      "grad_norm": 0.3472993075847626,
+      "learning_rate": 2.478065756904964e-05,
+      "loss": 0.9472295641899109,
+      "step": 112
+    },
+    {
+      "epoch": 0.4541832669322709,
+      "grad_norm": 0.4952745735645294,
+      "learning_rate": 2.476584280757419e-05,
+      "loss": 1.2006454467773438,
+      "step": 114
+    },
+    {
+      "epoch": 0.46215139442231074,
+      "grad_norm": 0.380439430475235,
+      "learning_rate": 2.4750549206462377e-05,
+      "loss": 1.2998346090316772,
+      "step": 116
+    },
+    {
+      "epoch": 0.4701195219123506,
+      "grad_norm": 0.39273232221603394,
+      "learning_rate": 2.4734777430500595e-05,
+      "loss": 1.3684892654418945,
+      "step": 118
+    },
+    {
+      "epoch": 0.47808764940239046,
+      "grad_norm": 0.3348349928855896,
+      "learning_rate": 2.4718528165260672e-05,
+      "loss": 1.2450612783432007,
+      "step": 120
+    },
+    {
+      "epoch": 0.4860557768924303,
+      "grad_norm": 0.788051962852478,
+      "learning_rate": 2.4701802117070067e-05,
+      "loss": 1.0636615753173828,
+      "step": 122
+    },
+    {
+      "epoch": 0.4940239043824701,
+      "grad_norm": 1.3081300258636475,
+      "learning_rate": 2.4684600012981167e-05,
+      "loss": 1.0938397645950317,
+      "step": 124
+    },
+    {
+      "epoch": 0.50199203187251,
+      "grad_norm": 2.2981116771698,
+      "learning_rate": 2.4666922600739686e-05,
+      "loss": 0.967792809009552,
+      "step": 126
+    },
+    {
+      "epoch": 0.5099601593625498,
+      "grad_norm": 0.32456639409065247,
+      "learning_rate": 2.464877064875217e-05,
+      "loss": 1.3107523918151855,
+      "step": 128
+    },
+    {
+      "epoch": 0.5179282868525896,
+      "grad_norm": 0.3772982954978943,
+      "learning_rate": 2.463014494605258e-05,
+      "loss": 0.9880162477493286,
+      "step": 130
+    },
+    {
+      "epoch": 0.5258964143426295,
+      "grad_norm": 0.4235079288482666,
+      "learning_rate": 2.4611046302267994e-05,
+      "loss": 0.9253523945808411,
+      "step": 132
+    },
+    {
+      "epoch": 0.5338645418326693,
+      "grad_norm": 0.4231255054473877,
+      "learning_rate": 2.4591475547583437e-05,
+      "loss": 0.9504383206367493,
+      "step": 134
+    },
+    {
+      "epoch": 0.5418326693227091,
+      "grad_norm": 0.6800777912139893,
+      "learning_rate": 2.4571433532705763e-05,
+      "loss": 1.408877968788147,
+      "step": 136
+    },
+    {
+      "epoch": 0.549800796812749,
+      "grad_norm": 0.7125083208084106,
+      "learning_rate": 2.45509211288267e-05,
+      "loss": 1.0226997137069702,
+      "step": 138
+    },
+    {
+      "epoch": 0.5577689243027888,
+      "grad_norm": 1.0820659399032593,
+      "learning_rate": 2.452993922758498e-05,
+      "loss": 1.1290892362594604,
+      "step": 140
+    },
+    {
+      "epoch": 0.5657370517928287,
+      "grad_norm": 0.32195162773132324,
+      "learning_rate": 2.450848874102755e-05,
+      "loss": 1.0651096105575562,
+      "step": 142
+    },
+    {
+      "epoch": 0.5737051792828686,
+      "grad_norm": 0.42358705401420593,
+      "learning_rate": 2.4486570601569977e-05,
+      "loss": 1.2509379386901855,
+      "step": 144
+    },
+    {
+      "epoch": 0.5816733067729084,
+      "grad_norm": 0.49886709451675415,
+      "learning_rate": 2.446418576195587e-05,
+      "loss": 1.1880295276641846,
+      "step": 146
+    },
+    {
+      "epoch": 0.5896414342629482,
+      "grad_norm": 0.6109864115715027,
+      "learning_rate": 2.4441335195215505e-05,
+      "loss": 1.3319215774536133,
+      "step": 148
+    },
+    {
+      "epoch": 0.5976095617529881,
+      "grad_norm": 0.5601715445518494,
+      "learning_rate": 2.441801989462349e-05,
+      "loss": 1.0226577520370483,
+      "step": 150
+    },
+    {
+      "epoch": 0.6055776892430279,
+      "grad_norm": 0.495827853679657,
+      "learning_rate": 2.439424087365563e-05,
+      "loss": 1.0290615558624268,
+      "step": 152
+    },
+    {
+      "epoch": 0.6135458167330677,
+      "grad_norm": 0.7617877125740051,
+      "learning_rate": 2.4369999165944833e-05,
+      "loss": 1.1918120384216309,
+      "step": 154
+    },
+    {
+      "epoch": 0.6215139442231076,
+      "grad_norm": 4.383106231689453,
+      "learning_rate": 2.4345295825236227e-05,
+      "loss": 1.454586148262024,
+      "step": 156
+    },
+    {
+      "epoch": 0.6294820717131474,
+      "grad_norm": 1.0111958980560303,
+      "learning_rate": 2.4320131925341295e-05,
+      "loss": 1.2641454935073853,
+      "step": 158
+    },
+    {
+      "epoch": 0.6374501992031872,
+      "grad_norm": 0.3744335174560547,
+      "learning_rate": 2.429450856009125e-05,
+      "loss": 1.1201077699661255,
+      "step": 160
+    },
+    {
+      "epoch": 0.6454183266932271,
+      "grad_norm": 0.2399037629365921,
+      "learning_rate": 2.426842684328948e-05,
+      "loss": 0.8693634867668152,
+      "step": 162
+    },
+    {
+      "epoch": 0.6533864541832669,
+      "grad_norm": 0.3195260465145111,
+      "learning_rate": 2.4241887908663095e-05,
+      "loss": 1.2242621183395386,
+      "step": 164
+    },
+    {
+      "epoch": 0.6613545816733067,
+      "grad_norm": 0.3846421241760254,
+      "learning_rate": 2.4214892909813697e-05,
+      "loss": 0.9046586751937866,
+      "step": 166
+    },
+    {
+      "epoch": 0.6693227091633466,
+      "grad_norm": 0.4162372946739197,
+      "learning_rate": 2.4187443020167196e-05,
+      "loss": 1.2568726539611816,
+      "step": 168
+    },
+    {
+      "epoch": 0.6772908366533864,
+      "grad_norm": 4.051391124725342,
+      "learning_rate": 2.4159539432922827e-05,
+      "loss": 1.0550285577774048,
+      "step": 170
+    },
+    {
+      "epoch": 0.6852589641434262,
+      "grad_norm": 0.3953443467617035,
+      "learning_rate": 2.4131183361001265e-05,
+      "loss": 0.9443556666374207,
+      "step": 172
+    },
+    {
+      "epoch": 0.6932270916334662,
+      "grad_norm": 0.35415229201316833,
+      "learning_rate": 2.4102376036991927e-05,
+      "loss": 1.2645931243896484,
+      "step": 174
+    },
+    {
+      "epoch": 0.701195219123506,
+      "grad_norm": 1.9177939891815186,
+      "learning_rate": 2.407311871309937e-05,
+      "loss": 0.8933344483375549,
+      "step": 176
+    },
+    {
+      "epoch": 0.7091633466135459,
+      "grad_norm": 0.7749300599098206,
+      "learning_rate": 2.404341266108886e-05,
+      "loss": 1.0225528478622437,
+      "step": 178
+    },
+    {
+      "epoch": 0.7171314741035857,
+      "grad_norm": 0.9419974684715271,
+      "learning_rate": 2.4013259172231106e-05,
+      "loss": 1.1758071184158325,
+      "step": 180
+    },
+    {
+      "epoch": 0.7250996015936255,
+      "grad_norm": 0.22870370745658875,
+      "learning_rate": 2.3982659557246125e-05,
+      "loss": 0.5740191340446472,
+      "step": 182
+    },
+    {
+      "epoch": 0.7330677290836654,
+      "grad_norm": 0.319692999124527,
+      "learning_rate": 2.395161514624626e-05,
+      "loss": 1.4750192165374756,
+      "step": 184
+    },
+    {
+      "epoch": 0.7410358565737052,
+      "grad_norm": 0.7873232960700989,
+      "learning_rate": 2.3920127288678375e-05,
+      "loss": 1.350590705871582,
+      "step": 186
+    },
+    {
+      "epoch": 0.749003984063745,
+      "grad_norm": 0.4544673562049866,
+      "learning_rate": 2.3888197353265175e-05,
+      "loss": 1.1976053714752197,
+      "step": 188
+    },
+    {
+      "epoch": 0.7569721115537849,
+      "grad_norm": 0.7830284833908081,
+      "learning_rate": 2.385582672794574e-05,
+      "loss": 0.9791259765625,
+      "step": 190
+    },
+    {
+      "epoch": 0.7649402390438247,
+      "grad_norm": 0.6947764158248901,
+      "learning_rate": 2.3823016819815176e-05,
+      "loss": 1.3110116720199585,
+      "step": 192
+    },
+    {
+      "epoch": 0.7729083665338645,
+      "grad_norm": 0.5548290610313416,
+      "learning_rate": 2.378976905506344e-05,
+      "loss": 0.9962631464004517,
+      "step": 194
+    },
+    {
+      "epoch": 0.7808764940239044,
+      "grad_norm": 1.9333361387252808,
+      "learning_rate": 2.3756084878913367e-05,
+      "loss": 0.7299926280975342,
+      "step": 196
+    },
+    {
+      "epoch": 0.7888446215139442,
+      "grad_norm": 0.6085161566734314,
+      "learning_rate": 2.372196575555785e-05,
+      "loss": 0.8126811981201172,
+      "step": 198
+    },
+    {
+      "epoch": 0.796812749003984,
+      "grad_norm": 1.438844919204712,
+      "learning_rate": 2.3687413168096174e-05,
+      "loss": 1.1386101245880127,
+      "step": 200
+    },
+    {
+      "epoch": 0.8047808764940239,
+      "grad_norm": 0.176991268992424,
+      "learning_rate": 2.365242861846956e-05,
+      "loss": 1.0891478061676025,
+      "step": 202
+    },
+    {
+      "epoch": 0.8127490039840638,
+      "grad_norm": 0.6198769211769104,
+      "learning_rate": 2.361701362739588e-05,
+      "loss": 1.6056677103042603,
+      "step": 204
+    },
+    {
+      "epoch": 0.8207171314741036,
+      "grad_norm": 0.5717342495918274,
+      "learning_rate": 2.3581169734303547e-05,
+      "loss": 1.2705366611480713,
+      "step": 206
+    },
+    {
+      "epoch": 0.8286852589641435,
+      "grad_norm": 0.3315287232398987,
+      "learning_rate": 2.3544898497264617e-05,
+      "loss": 0.6427626013755798,
+      "step": 208
+    },
+    {
+      "epoch": 0.8366533864541833,
+      "grad_norm": 0.5869395136833191,
+      "learning_rate": 2.3508201492927025e-05,
+      "loss": 1.3021904230117798,
+      "step": 210
+    },
+    {
+      "epoch": 0.8446215139442231,
+      "grad_norm": 0.9014328718185425,
+      "learning_rate": 2.3471080316446093e-05,
+      "loss": 1.2499291896820068,
+      "step": 212
+    },
+    {
+      "epoch": 0.852589641434263,
+      "grad_norm": 0.39842960238456726,
+      "learning_rate": 2.3433536581415162e-05,
+      "loss": 0.9948866963386536,
+      "step": 214
+    },
+    {
+      "epoch": 0.8605577689243028,
+      "grad_norm": 1.6447372436523438,
+      "learning_rate": 2.3395571919795477e-05,
+      "loss": 1.4756721258163452,
+      "step": 216
+    },
+    {
+      "epoch": 0.8685258964143426,
+      "grad_norm": 0.4688930809497833,
+      "learning_rate": 2.335718798184521e-05,
+      "loss": 1.0991026163101196,
+      "step": 218
+    },
+    {
+      "epoch": 0.8764940239043825,
+      "grad_norm": 0.4185577630996704,
+      "learning_rate": 2.331838643604777e-05,
+      "loss": 1.2656034231185913,
+      "step": 220
+    },
+    {
+      "epoch": 0.8844621513944223,
+      "grad_norm": 0.5186870694160461,
+      "learning_rate": 2.3279168969039253e-05,
+      "loss": 1.530430555343628,
+      "step": 222
+    },
+    {
+      "epoch": 0.8924302788844621,
+      "grad_norm": 0.35704922676086426,
+      "learning_rate": 2.3239537285535126e-05,
+      "loss": 0.7939154505729675,
+      "step": 224
+    },
+    {
+      "epoch": 0.900398406374502,
+      "grad_norm": 0.2659592926502228,
+      "learning_rate": 2.319949310825613e-05,
+      "loss": 1.0852367877960205,
+      "step": 226
+    },
+    {
+      "epoch": 0.9083665338645418,
+      "grad_norm": 0.4553312659263611,
+      "learning_rate": 2.3159038177853402e-05,
+      "loss": 1.1500097513198853,
+      "step": 228
+    },
+    {
+      "epoch": 0.9163346613545816,
+      "grad_norm": 1.1017969846725464,
+      "learning_rate": 2.311817425283281e-05,
+      "loss": 1.2199984788894653,
+      "step": 230
+    },
+    {
+      "epoch": 0.9243027888446215,
+      "grad_norm": 0.45185157656669617,
+      "learning_rate": 2.30769031094785e-05,
+      "loss": 1.3661251068115234,
+      "step": 232
+    },
+    {
+      "epoch": 0.9322709163346613,
+      "grad_norm": 0.44691282510757446,
+      "learning_rate": 2.3035226541775707e-05,
+      "loss": 0.8792702555656433,
+      "step": 234
+    },
+    {
+      "epoch": 0.9402390438247012,
+      "grad_norm": 0.3235927224159241,
+      "learning_rate": 2.2993146361332747e-05,
+      "loss": 0.8648804426193237,
+      "step": 236
+    },
+    {
+      "epoch": 0.9482071713147411,
+      "grad_norm": 0.4361308217048645,
+      "learning_rate": 2.295066439730229e-05,
+      "loss": 0.6265156269073486,
+      "step": 238
+    },
+    {
+      "epoch": 0.9561752988047809,
+      "grad_norm": 1.0246118307113647,
+      "learning_rate": 2.290778249630185e-05,
+      "loss": 0.8974942564964294,
+      "step": 240
+    },
+    {
+      "epoch": 0.9641434262948207,
+      "grad_norm": 0.8501946330070496,
+      "learning_rate": 2.286450252233351e-05,
+      "loss": 1.1724505424499512,
+      "step": 242
+    },
+    {
+      "epoch": 0.9721115537848606,
+      "grad_norm": 0.4002952575683594,
+      "learning_rate": 2.2820826356702885e-05,
+      "loss": 1.2344274520874023,
+      "step": 244
+    },
+    {
+      "epoch": 0.9800796812749004,
+      "grad_norm": 0.7706432342529297,
+      "learning_rate": 2.2776755897937356e-05,
+      "loss": 1.0097391605377197,
+      "step": 246
+    },
+    {
+      "epoch": 0.9880478087649402,
+      "grad_norm": 0.33801567554473877,
+      "learning_rate": 2.2732293061703565e-05,
+      "loss": 1.0792330503463745,
+      "step": 248
+    },
+    {
+      "epoch": 0.9960159362549801,
+      "grad_norm": 2.0120701789855957,
+      "learning_rate": 2.2687439780724104e-05,
+      "loss": 1.0766074657440186,
+      "step": 250
+    },
+    {
+      "epoch": 1.00398406374502,
+      "grad_norm": 0.3848106563091278,
+      "learning_rate": 2.264219800469353e-05,
+      "loss": 0.9873009324073792,
+      "step": 252
+    },
+    {
+      "epoch": 1.0119521912350598,
+      "grad_norm": 0.36586278676986694,
+      "learning_rate": 2.2596569700193604e-05,
+      "loss": 0.6772463917732239,
+      "step": 254
+    },
+    {
+      "epoch": 1.0199203187250996,
+      "grad_norm": 0.49873465299606323,
+      "learning_rate": 2.255055685060782e-05,
+      "loss": 1.0441770553588867,
+      "step": 256
+    },
+    {
+      "epoch": 1.0278884462151394,
+      "grad_norm": 0.4037606120109558,
+      "learning_rate": 2.2504161456035167e-05,
+      "loss": 1.0247764587402344,
+      "step": 258
+    },
+    {
+      "epoch": 1.0358565737051793,
+      "grad_norm": 0.43454790115356445,
+      "learning_rate": 2.2457385533203227e-05,
+      "loss": 0.9123398661613464,
+      "step": 260
+    },
+    {
+      "epoch": 1.043824701195219,
+      "grad_norm": 0.5795056819915771,
+      "learning_rate": 2.2410231115380456e-05,
+      "loss": 0.9118765592575073,
+      "step": 262
+    },
+    {
+      "epoch": 1.051792828685259,
+      "grad_norm": 0.5023526549339294,
+      "learning_rate": 2.236270025228786e-05,
+      "loss": 1.0351191759109497,
+      "step": 264
+    },
+    {
+      "epoch": 1.0597609561752988,
+      "grad_norm": 0.3814879059791565,
+      "learning_rate": 2.2314795010009854e-05,
+      "loss": 0.647515058517456,
+      "step": 266
+    },
+    {
+      "epoch": 1.0677290836653386,
+      "grad_norm": 0.22013379633426666,
+      "learning_rate": 2.2266517470904478e-05,
+      "loss": 0.8579058647155762,
+      "step": 268
+    },
+    {
+      "epoch": 1.0756972111553784,
+      "grad_norm": 0.2437354475259781,
+      "learning_rate": 2.2217869733512865e-05,
+      "loss": 0.5969091057777405,
+      "step": 270
+    },
+    {
+      "epoch": 1.0836653386454183,
+      "grad_norm": 0.4381161630153656,
+      "learning_rate": 2.216885391246803e-05,
+      "loss": 1.1901330947875977,
+      "step": 272
+    },
+    {
+      "epoch": 1.091633466135458,
+      "grad_norm": 0.6550458669662476,
+      "learning_rate": 2.2119472138402948e-05,
+      "loss": 0.7100527882575989,
+      "step": 274
+    },
+    {
+      "epoch": 1.099601593625498,
+      "grad_norm": 0.3357766270637512,
+      "learning_rate": 2.206972655785793e-05,
+      "loss": 0.7771211266517639,
+      "step": 276
+    },
+    {
+      "epoch": 1.1075697211155378,
+      "grad_norm": 0.39857205748558044,
+      "learning_rate": 2.2019619333187346e-05,
+      "loss": 0.737285315990448,
+      "step": 278
+    },
+    {
+      "epoch": 1.1155378486055776,
+      "grad_norm": 1.949383020401001,
+      "learning_rate": 2.196915264246559e-05,
+      "loss": 1.027347207069397,
+      "step": 280
+    },
+    {
+      "epoch": 1.1235059760956174,
+      "grad_norm": 0.58866947889328,
+      "learning_rate": 2.1918328679392438e-05,
+      "loss": 1.0846821069717407,
+      "step": 282
+    },
+    {
+      "epoch": 1.1314741035856573,
+      "grad_norm": 0.5925541520118713,
+      "learning_rate": 2.186714965319767e-05,
+      "loss": 1.1215362548828125,
+      "step": 284
+    },
+    {
+      "epoch": 1.139442231075697,
+      "grad_norm": 1.182293176651001,
+      "learning_rate": 2.1815617788545043e-05,
+      "loss": 0.42030107975006104,
+      "step": 286
+    },
+    {
+      "epoch": 1.1474103585657371,
+      "grad_norm": 0.1760035902261734,
+      "learning_rate": 2.1763735325435597e-05,
+      "loss": 0.6199245452880859,
+      "step": 288
+    },
+    {
+      "epoch": 1.155378486055777,
+      "grad_norm": 0.8143728375434875,
+      "learning_rate": 2.1711504519110282e-05,
+      "loss": 0.8853223323822021,
+      "step": 290
+    },
+    {
+      "epoch": 1.1633466135458168,
+      "grad_norm": 0.44977709650993347,
+      "learning_rate": 2.165892763995192e-05,
+      "loss": 0.5433615446090698,
+      "step": 292
+    },
+    {
+      "epoch": 1.1713147410358566,
+      "grad_norm": 0.2932179868221283,
+      "learning_rate": 2.160600697338652e-05,
+      "loss": 1.1232950687408447,
+      "step": 294
+    },
+    {
+      "epoch": 1.1792828685258965,
+      "grad_norm": 0.3169727921485901,
+      "learning_rate": 2.1552744819783953e-05,
+      "loss": 1.1833813190460205,
+      "step": 296
+    },
+    {
+      "epoch": 1.1872509960159363,
+      "grad_norm": 0.40002286434173584,
+      "learning_rate": 2.1499143494357905e-05,
+      "loss": 1.1253842115402222,
+      "step": 298
+    },
+    {
+      "epoch": 1.1952191235059761,
+      "grad_norm": 0.7417656183242798,
+      "learning_rate": 2.144520532706531e-05,
+      "loss": 0.6202850937843323,
+      "step": 300
+    },
+    {
+      "epoch": 1.203187250996016,
+      "grad_norm": 0.4357120990753174,
+      "learning_rate": 2.1390932662505013e-05,
+      "loss": 0.8437725901603699,
+      "step": 302
+    },
+    {
+      "epoch": 1.2111553784860558,
+      "grad_norm": 0.43674275279045105,
+      "learning_rate": 2.1336327859815892e-05,
+      "loss": 0.5135228037834167,
+      "step": 304
+    },
+    {
+      "epoch": 1.2191235059760956,
+      "grad_norm": 0.532661497592926,
+      "learning_rate": 2.1281393292574287e-05,
+      "loss": 1.096781849861145,
+      "step": 306
+    },
+    {
+      "epoch": 1.2270916334661355,
+      "grad_norm": 0.8392455577850342,
+      "learning_rate": 2.1226131348690828e-05,
+      "loss": 0.8513472080230713,
+      "step": 308
+    },
+    {
+      "epoch": 1.2350597609561753,
+      "grad_norm": 0.7447803616523743,
+      "learning_rate": 2.1170544430306653e-05,
+      "loss": 1.0444648265838623,
+      "step": 310
+    },
+    {
+      "epoch": 1.2430278884462151,
+      "grad_norm": 0.4204055964946747,
+      "learning_rate": 2.1114634953688967e-05,
+      "loss": 0.9815379977226257,
+      "step": 312
+    },
+    {
+      "epoch": 1.250996015936255,
+      "grad_norm": 0.5915699601173401,
+      "learning_rate": 2.1058405349126034e-05,
+      "loss": 0.844771683216095,
+      "step": 314
+    },
+    {
+      "epoch": 1.2589641434262948,
+      "grad_norm": 0.8848176598548889,
+      "learning_rate": 2.1001858060821527e-05,
+      "loss": 0.7396162152290344,
+      "step": 316
+    },
+    {
+      "epoch": 1.2669322709163346,
+      "grad_norm": 0.782599687576294,
+      "learning_rate": 2.0944995546788275e-05,
+      "loss": 0.6623770594596863,
+      "step": 318
+    },
+    {
+      "epoch": 1.2749003984063745,
+      "grad_norm": 0.566226065158844,
+      "learning_rate": 2.088782027874142e-05,
+      "loss": 0.6006486415863037,
+      "step": 320
+    },
+    {
+      "epoch": 1.2828685258964143,
+      "grad_norm": 1.0356029272079468,
+      "learning_rate": 2.0830334741990995e-05,
+      "loss": 0.9051658511161804,
+      "step": 322
+    },
+    {
+      "epoch": 1.2908366533864541,
+      "grad_norm": 0.386123925447464,
+      "learning_rate": 2.077254143533386e-05,
+      "loss": 0.8423581719398499,
+      "step": 324
+    },
+    {
+      "epoch": 1.298804780876494,
+      "grad_norm": 0.9391600489616394,
+      "learning_rate": 2.0714442870945122e-05,
+      "loss": 1.2191598415374756,
+      "step": 326
+    },
+    {
+      "epoch": 1.3067729083665338,
+      "grad_norm": 0.22411197423934937,
+      "learning_rate": 2.065604157426889e-05,
+      "loss": 0.6406465172767639,
+      "step": 328
+    },
+    {
+      "epoch": 1.3147410358565736,
+      "grad_norm": 0.40532219409942627,
+      "learning_rate": 2.0597340083908537e-05,
+      "loss": 0.6966909766197205,
+      "step": 330
+    },
+    {
+      "epoch": 1.3227091633466135,
+      "grad_norm": 0.5293474197387695,
+      "learning_rate": 2.053834095151633e-05,
+      "loss": 0.8643190860748291,
+      "step": 332
+    },
+    {
+      "epoch": 1.3306772908366533,
+      "grad_norm": 0.7702229022979736,
+      "learning_rate": 2.0479046741682527e-05,
+      "loss": 0.6109381914138794,
+      "step": 334
+    },
+    {
+      "epoch": 1.3386454183266931,
+      "grad_norm": 0.6019669771194458,
+      "learning_rate": 2.0419460031823886e-05,
+      "loss": 0.8153576254844666,
+      "step": 336
+    },
+    {
+      "epoch": 1.3466135458167332,
+      "grad_norm": 0.5080145001411438,
+      "learning_rate": 2.0359583412071636e-05,
+      "loss": 1.0472195148468018,
+      "step": 338
+    },
+    {
+      "epoch": 1.354581673306773,
+      "grad_norm": 0.2602000832557678,
+      "learning_rate": 2.0299419485158882e-05,
+      "loss": 0.990772545337677,
+      "step": 340
+    },
+    {
+      "epoch": 1.3625498007968129,
+      "grad_norm": 0.20945246517658234,
+      "learning_rate": 2.0238970866307496e-05,
+      "loss": 1.1070572137832642,
+      "step": 342
+    },
+    {
+      "epoch": 1.3705179282868527,
+      "grad_norm": 0.31776490807533264,
+      "learning_rate": 2.0178240183114393e-05,
+      "loss": 1.0254907608032227,
+      "step": 344
+    },
+    {
+      "epoch": 1.3784860557768925,
+      "grad_norm": 0.6745681166648865,
+      "learning_rate": 2.0117230075437344e-05,
+      "loss": 0.6767964363098145,
+      "step": 346
+    },
+    {
+      "epoch": 1.3864541832669324,
+      "grad_norm": 0.250434547662735,
+      "learning_rate": 2.005594319528022e-05,
+      "loss": 1.0467485189437866,
+      "step": 348
+    },
+    {
+      "epoch": 1.3944223107569722,
+      "grad_norm": 0.2523025870323181,
+      "learning_rate": 1.999438220667772e-05,
+      "loss": 0.617721676826477,
+      "step": 350
+    },
+    {
+      "epoch": 1.402390438247012,
+      "grad_norm": 0.25256848335266113,
+      "learning_rate": 1.9932549785579548e-05,
+      "loss": 1.072009563446045,
+      "step": 352
+    },
+    {
+      "epoch": 1.4103585657370519,
+      "grad_norm": 0.3956409990787506,
+      "learning_rate": 1.987044861973413e-05,
+      "loss": 1.0613605976104736,
+      "step": 354
+    },
+    {
+      "epoch": 1.4183266932270917,
+      "grad_norm": 0.32173997163772583,
+      "learning_rate": 1.980808140857173e-05,
+      "loss": 0.8194036483764648,
+      "step": 356
+    },
+    {
+      "epoch": 1.4262948207171315,
+      "grad_norm": 0.9387481808662415,
+      "learning_rate": 1.9745450863087185e-05,
+      "loss": 1.1056541204452515,
+      "step": 358
+    },
+    {
+      "epoch": 1.4342629482071714,
+      "grad_norm": 0.5239622592926025,
+      "learning_rate": 1.9682559705721985e-05,
+      "loss": 0.5867342352867126,
+      "step": 360
+    },
+    {
+      "epoch": 1.4422310756972112,
+      "grad_norm": 0.32403695583343506,
+      "learning_rate": 1.9619410670245968e-05,
+      "loss": 0.6357664465904236,
+      "step": 362
+    },
+    {
+      "epoch": 1.450199203187251,
+      "grad_norm": 0.37655097246170044,
+      "learning_rate": 1.9556006501638512e-05,
+      "loss": 1.0264805555343628,
+      "step": 364
+    },
+    {
+      "epoch": 1.4581673306772909,
+      "grad_norm": 0.26585426926612854,
+      "learning_rate": 1.9492349955969174e-05,
+      "loss": 1.0221697092056274,
+      "step": 366
+    },
+    {
+      "epoch": 1.4661354581673307,
+      "grad_norm": 0.7691898941993713,
+      "learning_rate": 1.9428443800277907e-05,
+      "loss": 0.5706173181533813,
+      "step": 368
+    },
+    {
+      "epoch": 1.4741035856573705,
+      "grad_norm": 0.413251131772995,
+      "learning_rate": 1.9364290812454778e-05,
+      "loss": 0.6572546362876892,
+      "step": 370
+    },
+    {
+      "epoch": 1.4820717131474104,
+      "grad_norm": 0.8349089622497559,
+      "learning_rate": 1.9299893781119227e-05,
+      "loss": 1.2169642448425293,
+      "step": 372
+    },
+    {
+      "epoch": 1.4900398406374502,
+      "grad_norm": 0.2802586853504181,
+      "learning_rate": 1.9235255505498843e-05,
+      "loss": 0.9851334691047668,
+      "step": 374
+    },
+    {
+      "epoch": 1.49800796812749,
+      "grad_norm": 0.22245906293392181,
+      "learning_rate": 1.9170378795307676e-05,
+      "loss": 1.0062472820281982,
+      "step": 376
+    },
+    {
+      "epoch": 1.5059760956175299,
+      "grad_norm": 1.4086861610412598,
+      "learning_rate": 1.9105266470624128e-05,
+      "loss": 0.8641257882118225,
+      "step": 378
+    },
+    {
+      "epoch": 1.5139442231075697,
+      "grad_norm": 0.335869699716568,
+      "learning_rate": 1.9039921361768342e-05,
+      "loss": 1.0744060277938843,
+      "step": 380
+    },
+    {
+      "epoch": 1.5219123505976095,
+      "grad_norm": 0.27242788672447205,
+      "learning_rate": 1.897434630917921e-05,
+      "loss": 0.6748390197753906,
+      "step": 382
+    },
+    {
+      "epoch": 1.5298804780876494,
+      "grad_norm": 0.34286174178123474,
+      "learning_rate": 1.8908544163290846e-05,
+      "loss": 0.6672536134719849,
+      "step": 384
+    },
+    {
+      "epoch": 1.5378486055776892,
+      "grad_norm": 1.9025431871414185,
+      "learning_rate": 1.8842517784408743e-05,
+      "loss": 0.9685118794441223,
+      "step": 386
+    },
+    {
+      "epoch": 1.545816733067729,
+      "grad_norm": 0.5212134122848511,
+      "learning_rate": 1.877627004258541e-05,
+      "loss": 1.0655114650726318,
+      "step": 388
+    },
+    {
+      "epoch": 1.5537848605577689,
+      "grad_norm": 0.5611221194267273,
+      "learning_rate": 1.8709803817495617e-05,
+      "loss": 0.6233720183372498,
+      "step": 390
+    },
+    {
+      "epoch": 1.5617529880478087,
+      "grad_norm": 0.7503693103790283,
+      "learning_rate": 1.8643121998311218e-05,
+      "loss": 0.542980432510376,
+      "step": 392
+    },
+    {
+      "epoch": 1.5697211155378485,
+      "grad_norm": 0.5970079302787781,
+      "learning_rate": 1.857622748357558e-05,
+      "loss": 1.1312108039855957,
+      "step": 394
+    },
+    {
+      "epoch": 1.5776892430278884,
+      "grad_norm": 0.661026656627655,
+      "learning_rate": 1.8509123181077567e-05,
+      "loss": 0.9834690690040588,
+      "step": 396
+    },
+    {
+      "epoch": 1.5856573705179282,
+      "grad_norm": 0.21081377565860748,
+      "learning_rate": 1.844181200772517e-05,
+      "loss": 1.0674799680709839,
+      "step": 398
+    },
+    {
+      "epoch": 1.593625498007968,
+      "grad_norm": 0.33451369404792786,
+      "learning_rate": 1.8374296889418683e-05,
+      "loss": 0.5782875418663025,
+      "step": 400
+    },
+    {
+      "epoch": 1.6015936254980079,
+      "grad_norm": 0.490013062953949,
+      "learning_rate": 1.8306580760923538e-05,
+      "loss": 1.0008902549743652,
+      "step": 402
+    },
+    {
+      "epoch": 1.6095617529880477,
+      "grad_norm": 0.2503955662250519,
+      "learning_rate": 1.823866656574274e-05,
+      "loss": 1.1085338592529297,
+      "step": 404
+    },
+    {
+      "epoch": 1.6175298804780875,
+      "grad_norm": 0.9677779078483582,
+      "learning_rate": 1.8170557255988906e-05,
+      "loss": 0.7185835242271423,
+      "step": 406
+    },
+    {
+      "epoch": 1.6254980079681274,
+      "grad_norm": 0.33572229743003845,
+      "learning_rate": 1.810225579225593e-05,
+      "loss": 0.5569627285003662,
+      "step": 408
+    },
+    {
+      "epoch": 1.6334661354581672,
+      "grad_norm": 0.26946014165878296,
+      "learning_rate": 1.8033765143490337e-05,
+      "loss": 0.6784675717353821,
+      "step": 410
+    },
+    {
+      "epoch": 1.641434262948207,
+      "grad_norm": 0.482138454914093,
+      "learning_rate": 1.7965088286862177e-05,
+      "loss": 0.48974257707595825,
+      "step": 412
+    },
+    {
+      "epoch": 1.6494023904382469,
+      "grad_norm": 0.2134891152381897,
+      "learning_rate": 1.7896228207635643e-05,
+      "loss": 0.7183177471160889,
+      "step": 414
+    },
+    {
+      "epoch": 1.6573705179282867,
+      "grad_norm": 0.39308246970176697,
+      "learning_rate": 1.782718789903929e-05,
+      "loss": 0.7660385370254517,
+      "step": 416
+    },
+    {
+      "epoch": 1.6653386454183265,
+      "grad_norm": 0.41467955708503723,
+      "learning_rate": 1.775797036213594e-05,
+      "loss": 0.8051981925964355,
+      "step": 418
+    },
+    {
+      "epoch": 1.6733067729083664,
+      "grad_norm": 0.3546353280544281,
+      "learning_rate": 1.768857860569222e-05,
+      "loss": 1.0917823314666748,
+      "step": 420
+    },
+    {
+      "epoch": 1.6812749003984062,
+      "grad_norm": 0.2528360188007355,
+      "learning_rate": 1.761901564604777e-05,
+      "loss": 0.9082453846931458,
+      "step": 422
+    },
+    {
+      "epoch": 1.6892430278884463,
+      "grad_norm": 0.5713571906089783,
+      "learning_rate": 1.7549284506984155e-05,
+      "loss": 0.6437570452690125,
+      "step": 424
+    },
+    {
+      "epoch": 1.697211155378486,
+      "grad_norm": 0.2368750423192978,
+      "learning_rate": 1.7479388219593398e-05,
+      "loss": 1.11428964138031,
+      "step": 426
+    },
+    {
+      "epoch": 1.705179282868526,
+      "grad_norm": 0.36424264311790466,
+      "learning_rate": 1.7409329822146237e-05,
+      "loss": 1.1256872415542603,
+      "step": 428
+    },
+    {
+      "epoch": 1.7131474103585658,
+      "grad_norm": 0.2989889979362488,
+      "learning_rate": 1.7339112359960042e-05,
+      "loss": 0.7030774354934692,
+      "step": 430
+    },
+    {
+      "epoch": 1.7211155378486056,
+      "grad_norm": 0.8977339863777161,
+      "learning_rate": 1.7268738885266466e-05,
+      "loss": 0.840191662311554,
+      "step": 432
+    },
+    {
+      "epoch": 1.7290836653386454,
+      "grad_norm": 0.5040406584739685,
+      "learning_rate": 1.7198212457078762e-05,
+      "loss": 0.6149595379829407,
+      "step": 434
+    },
+    {
+      "epoch": 1.7370517928286853,
+      "grad_norm": 0.902721643447876,
+      "learning_rate": 1.7127536141058785e-05,
+      "loss": 0.42500048875808716,
+      "step": 436
+    },
+    {
+      "epoch": 1.745019920318725,
+      "grad_norm": 0.6491820812225342,
+      "learning_rate": 1.705671300938377e-05,
+      "loss": 0.7307272553443909,
+      "step": 438
+    },
+    {
+      "epoch": 1.752988047808765,
+      "grad_norm": 3.873206853866577,
+      "learning_rate": 1.6985746140612777e-05,
+      "loss": 0.6927053928375244,
+      "step": 440
+    },
+    {
+      "epoch": 1.7609561752988048,
+      "grad_norm": 0.6419493556022644,
+      "learning_rate": 1.6914638619552857e-05,
+      "loss": 1.3595691919326782,
+      "step": 442
+    },
+    {
+      "epoch": 1.7689243027888446,
+      "grad_norm": 0.6623972058296204,
+      "learning_rate": 1.684339353712498e-05,
+      "loss": 0.7792170643806458,
+      "step": 444
+    },
+    {
+      "epoch": 1.7768924302788844,
+      "grad_norm": 0.3009021580219269,
+      "learning_rate": 1.6772013990229674e-05,
+      "loss": 0.9983497262001038,
+      "step": 446
+    },
+    {
+      "epoch": 1.7848605577689243,
+      "grad_norm": 0.8442715406417847,
+      "learning_rate": 1.6700503081612396e-05,
+      "loss": 1.011168122291565,
+      "step": 448
+    },
+    {
+      "epoch": 1.792828685258964,
+      "grad_norm": 0.34507128596305847,
+      "learning_rate": 1.6628863919728677e-05,
+      "loss": 0.8132411241531372,
+      "step": 450
+    },
+    {
+      "epoch": 1.800796812749004,
+      "grad_norm": 0.16587740182876587,
+      "learning_rate": 1.655709961860899e-05,
+      "loss": 0.4028864800930023,
+      "step": 452
+    },
+    {
+      "epoch": 1.8087649402390438,
+      "grad_norm": 0.39269912242889404,
+      "learning_rate": 1.6485213297723402e-05,
+      "loss": 0.8867176175117493,
+      "step": 454
+    },
+    {
+      "epoch": 1.8167330677290838,
+      "grad_norm": 0.5646684765815735,
+      "learning_rate": 1.6413208081845964e-05,
+      "loss": 0.6442360281944275,
+      "step": 456
+    },
+    {
+      "epoch": 1.8247011952191237,
+      "grad_norm": 1.0004042387008667,
+      "learning_rate": 1.6341087100918886e-05,
+      "loss": 1.0023376941680908,
+      "step": 458
+    },
+    {
+      "epoch": 1.8326693227091635,
+      "grad_norm": 0.2544229328632355,
+      "learning_rate": 1.6268853489916478e-05,
+      "loss": 0.7372154593467712,
+      "step": 460
+    },
+    {
+      "epoch": 1.8406374501992033,
+      "grad_norm": 0.32100027799606323,
+      "learning_rate": 1.61965103887089e-05,
+      "loss": 0.7921067476272583,
+      "step": 462
+    },
+    {
+      "epoch": 1.8486055776892432,
+      "grad_norm": 0.21843624114990234,
+      "learning_rate": 1.6124060941925646e-05,
+      "loss": 0.14675641059875488,
+      "step": 464
+    },
+    {
+      "epoch": 1.856573705179283,
+      "grad_norm": 0.34702828526496887,
+      "learning_rate": 1.6051508298818884e-05,
+      "loss": 1.0574944019317627,
+      "step": 466
+    },
+    {
+      "epoch": 1.8645418326693228,
+      "grad_norm": 0.49059736728668213,
+      "learning_rate": 1.5978855613126538e-05,
+      "loss": 0.49086442589759827,
+      "step": 468
+    },
+    {
+      "epoch": 1.8725099601593627,
+      "grad_norm": 0.46234646439552307,
+      "learning_rate": 1.5906106042935205e-05,
+      "loss": 1.1290267705917358,
+      "step": 470
+    },
+    {
+      "epoch": 1.8804780876494025,
+      "grad_norm": 0.6888619661331177,
+      "learning_rate": 1.58332627505429e-05,
+      "loss": 0.8859848976135254,
+      "step": 472
+    },
+    {
+      "epoch": 1.8884462151394423,
+      "grad_norm": 1.7080014944076538,
+      "learning_rate": 1.5760328902321555e-05,
+      "loss": 1.154191255569458,
+      "step": 474
+    },
+    {
+      "epoch": 1.8964143426294822,
+      "grad_norm": 0.3313632607460022,
+      "learning_rate": 1.5687307668579437e-05,
+      "loss": 0.9348896741867065,
+      "step": 476
+    },
+    {
+      "epoch": 1.904382470119522,
+      "grad_norm": 0.25657084584236145,
+      "learning_rate": 1.561420222342329e-05,
+      "loss": 0.786576509475708,
+      "step": 478
+    },
+    {
+      "epoch": 1.9123505976095618,
+      "grad_norm": 3.075324535369873,
+      "learning_rate": 1.554101574462039e-05,
+      "loss": 0.5916264653205872,
+      "step": 480
+    },
+    {
+      "epoch": 1.9203187250996017,
+      "grad_norm": 0.619073212146759,
+      "learning_rate": 1.5467751413460404e-05,
+      "loss": 0.7957924008369446,
+      "step": 482
+    },
+    {
+      "epoch": 1.9282868525896415,
+      "grad_norm": 0.2277441918849945,
+      "learning_rate": 1.53944124146171e-05,
+      "loss": 1.013961672782898,
+      "step": 484
+    },
+    {
+      "epoch": 1.9362549800796813,
+      "grad_norm": 0.34578096866607666,
+      "learning_rate": 1.5321001936009934e-05,
+      "loss": 0.913303792476654,
+      "step": 486
+    },
+    {
+      "epoch": 1.9442231075697212,
+      "grad_norm": 0.24846151471138,
+      "learning_rate": 1.524752316866546e-05,
+      "loss": 0.9955238103866577,
+      "step": 488
+    },
+    {
+      "epoch": 1.952191235059761,
+      "grad_norm": 0.3897765576839447,
+      "learning_rate": 1.5173979306578615e-05,
+      "loss": 1.0086604356765747,
+      "step": 490
+    },
+    {
+      "epoch": 1.9601593625498008,
+      "grad_norm": 0.2907620668411255,
+      "learning_rate": 1.5100373546573904e-05,
+      "loss": 0.7454448342323303,
+      "step": 492
+    },
+    {
+      "epoch": 1.9681274900398407,
+      "grad_norm": 5.719922065734863,
+      "learning_rate": 1.5026709088166422e-05,
+      "loss": 0.76938796043396,
+      "step": 494
+    },
+    {
+      "epoch": 1.9760956175298805,
+      "grad_norm": 0.3191843032836914,
+      "learning_rate": 1.4952989133422782e-05,
+      "loss": 0.9833912253379822,
+      "step": 496
+    },
+    {
+      "epoch": 1.9840637450199203,
+      "grad_norm": 3.906707763671875,
+      "learning_rate": 1.4879216886821925e-05,
+      "loss": 0.7836262583732605,
+      "step": 498
+    },
+    {
+      "epoch": 1.9920318725099602,
+      "grad_norm": 0.23314878344535828,
+      "learning_rate": 1.4805395555115834e-05,
+      "loss": 1.1056411266326904,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.4651053845882416,
+      "learning_rate": 1.4731528347190133e-05,
+      "loss": 0.5553019046783447,
+      "step": 502
+    },
+    {
+      "epoch": 2.00796812749004,
+      "grad_norm": 0.3860473334789276,
+      "learning_rate": 1.4657618473924606e-05,
+      "loss": 0.44113031029701233,
+      "step": 504
+    },
+    {
+      "epoch": 2.0159362549800797,
+      "grad_norm": 0.40843892097473145,
+      "learning_rate": 1.458366914805363e-05,
+      "loss": 0.4357832968235016,
+      "step": 506
+    },
+    {
+      "epoch": 2.0239043824701195,
+      "grad_norm": 0.1141304150223732,
+      "learning_rate": 1.450968358402651e-05,
+      "loss": 0.3898138999938965,
+      "step": 508
+    },
+    {
+      "epoch": 2.0318725099601593,
+      "grad_norm": 0.23500245809555054,
+      "learning_rate": 1.443566499786777e-05,
+      "loss": 0.6758836507797241,
+      "step": 510
+    },
+    {
+      "epoch": 2.039840637450199,
+      "grad_norm": 0.4461332857608795,
+      "learning_rate": 1.4361616607037338e-05,
+      "loss": 0.5746989250183105,
+      "step": 512
+    },
+    {
+      "epoch": 2.047808764940239,
+      "grad_norm": 4.716413497924805,
+      "learning_rate": 1.428754163029071e-05,
+      "loss": 0.3819912075996399,
+      "step": 514
+    },
+    {
+      "epoch": 2.055776892430279,
+      "grad_norm": 0.3279768228530884,
+      "learning_rate": 1.4213443287539025e-05,
+      "loss": 0.6755325794219971,
+      "step": 516
+    },
+    {
+      "epoch": 2.0637450199203187,
+      "grad_norm": 1.3221046924591064,
+      "learning_rate": 1.4139324799709097e-05,
+      "loss": 0.4200603663921356,
+      "step": 518
+    },
+    {
+      "epoch": 2.0717131474103585,
+      "grad_norm": 0.5988231301307678,
+      "learning_rate": 1.4065189388603414e-05,
+      "loss": 0.5700657367706299,
+      "step": 520
+    },
+    {
+      "epoch": 2.0796812749003983,
+      "grad_norm": 0.7151240706443787,
+      "learning_rate": 1.3991040276760084e-05,
+      "loss": 0.699086606502533,
+      "step": 522
+    },
+    {
+      "epoch": 2.087649402390438,
+      "grad_norm": 0.3767349421977997,
+      "learning_rate": 1.3916880687312772e-05,
+      "loss": 0.7041755318641663,
+      "step": 524
+    },
+    {
+      "epoch": 2.095617529880478,
+      "grad_norm": 21.821245193481445,
+      "learning_rate": 1.3842713843850586e-05,
+      "loss": 0.32073432207107544,
+      "step": 526
+    },
+    {
+      "epoch": 2.103585657370518,
+      "grad_norm": 0.3366837501525879,
+      "learning_rate": 1.3768542970277939e-05,
+      "loss": 0.537539005279541,
+      "step": 528
+    },
+    {
+      "epoch": 2.1115537848605577,
+      "grad_norm": 0.14259065687656403,
+      "learning_rate": 1.3694371290674446e-05,
+      "loss": 0.4907974302768707,
+      "step": 530
+    },
+    {
+      "epoch": 2.1195219123505975,
+      "grad_norm": 0.562319815158844,
+      "learning_rate": 1.3620202029154745e-05,
+      "loss": 0.19045501947402954,
+      "step": 532
+    },
+    {
+      "epoch": 2.1274900398406373,
+      "grad_norm": 0.5693976879119873,
+      "learning_rate": 1.3546038409728366e-05,
+      "loss": 0.5350707769393921,
+      "step": 534
+    },
+    {
+      "epoch": 2.135458167330677,
+      "grad_norm": 0.7718969583511353,
+      "learning_rate": 1.3471883656159584e-05,
+      "loss": 0.5544182658195496,
+      "step": 536
+    },
+    {
+      "epoch": 2.143426294820717,
+      "grad_norm": 1.18233060836792,
+      "learning_rate": 1.3397740991827309e-05,
+      "loss": 0.5668792128562927,
+      "step": 538
+    },
+    {
+      "epoch": 2.151394422310757,
+      "grad_norm": 0.3552509844303131,
+      "learning_rate": 1.3323613639584937e-05,
+      "loss": 0.6888652443885803,
+      "step": 540
+    },
+    {
+      "epoch": 2.1593625498007967,
+      "grad_norm": 0.29430294036865234,
+      "learning_rate": 1.3249504821620261e-05,
+      "loss": 0.6434365510940552,
+      "step": 542
+    },
+    {
+      "epoch": 2.1673306772908365,
+      "grad_norm": 0.29388901591300964,
+      "learning_rate": 1.3175417759315453e-05,
+      "loss": 0.4334697127342224,
+      "step": 544
+    },
+    {
+      "epoch": 2.1752988047808763,
+      "grad_norm": 0.743347704410553,
+      "learning_rate": 1.3101355673106985e-05,
+      "loss": 0.4870750904083252,
+      "step": 546
+    },
+    {
+      "epoch": 2.183266932270916,
+      "grad_norm": 0.671739399433136,
+      "learning_rate": 1.3027321782345655e-05,
+      "loss": 0.6575522422790527,
+      "step": 548
+    },
+    {
+      "epoch": 2.191235059760956,
+      "grad_norm": 0.6698936223983765,
+      "learning_rate": 1.2953319305156666e-05,
+      "loss": 0.8323569297790527,
+      "step": 550
+    },
+    {
+      "epoch": 2.199203187250996,
+      "grad_norm": 0.7439915537834167,
+      "learning_rate": 1.2879351458299729e-05,
+      "loss": 0.4803932309150696,
+      "step": 552
+    },
+    {
+      "epoch": 2.2071713147410357,
+      "grad_norm": 0.7964186668395996,
+      "learning_rate": 1.2805421457029224e-05,
+      "loss": 0.4710814356803894,
+      "step": 554
+    },
+    {
+      "epoch": 2.2151394422310755,
+      "grad_norm": 0.2705269157886505,
+      "learning_rate": 1.2731532514954454e-05,
+      "loss": 0.5913985967636108,
+      "step": 556
+    },
+    {
+      "epoch": 2.2231075697211153,
+      "grad_norm": 0.6730039119720459,
+      "learning_rate": 1.2657687843899949e-05,
+      "loss": 0.7226806282997131,
+      "step": 558
+    },
+    {
+      "epoch": 2.231075697211155,
+      "grad_norm": 0.24639157950878143,
+      "learning_rate": 1.2583890653765856e-05,
+      "loss": 0.7639114856719971,
+      "step": 560
+    },
+    {
+      "epoch": 2.239043824701195,
+      "grad_norm": 0.3987577557563782,
+      "learning_rate": 1.2510144152388411e-05,
+      "loss": 0.7780134081840515,
+      "step": 562
+    },
+    {
+      "epoch": 2.247011952191235,
+      "grad_norm": 0.6101713180541992,
+      "learning_rate": 1.2436451545400484e-05,
+      "loss": 0.6677888035774231,
+      "step": 564
+    },
+    {
+      "epoch": 2.2549800796812747,
+      "grad_norm": 0.27706581354141235,
+      "learning_rate": 1.236281603609227e-05,
+      "loss": 0.40493103861808777,
+      "step": 566
+    },
+    {
+      "epoch": 2.2629482071713145,
+      "grad_norm": 0.22308364510536194,
+      "learning_rate": 1.228924082527201e-05,
+      "loss": 0.5804494023323059,
+      "step": 568
+    },
+    {
+      "epoch": 2.2709163346613543,
+      "grad_norm": 0.43112289905548096,
+      "learning_rate": 1.2215729111126877e-05,
+      "loss": 0.633328378200531,
+      "step": 570
+    },
+    {
+      "epoch": 2.278884462151394,
+      "grad_norm": 0.7963998317718506,
+      "learning_rate": 1.2142284089083961e-05,
+      "loss": 0.5453795790672302,
+      "step": 572
+    },
+    {
+      "epoch": 2.2868525896414345,
+      "grad_norm": 0.590513288974762,
+      "learning_rate": 1.206890895167136e-05,
+      "loss": 0.558796763420105,
+      "step": 574
+    },
+    {
+      "epoch": 2.2948207171314743,
+      "grad_norm": 0.6780595779418945,
+      "learning_rate": 1.1995606888379387e-05,
+      "loss": 0.23361682891845703,
+      "step": 576
+    },
+    {
+      "epoch": 2.302788844621514,
+      "grad_norm": 0.4337788224220276,
+      "learning_rate": 1.1922381085521985e-05,
+      "loss": 0.4590423107147217,
+      "step": 578
+    },
+    {
+      "epoch": 2.310756972111554,
+      "grad_norm": 0.4065179228782654,
+      "learning_rate": 1.1849234726098163e-05,
+      "loss": 0.447817325592041,
+      "step": 580
+    },
+    {
+      "epoch": 2.318725099601594,
+      "grad_norm": 3.771479845046997,
+      "learning_rate": 1.1776170989653674e-05,
+      "loss": 0.6217383742332458,
+      "step": 582
+    },
+    {
+      "epoch": 2.3266932270916336,
+      "grad_norm": 0.3368018865585327,
+      "learning_rate": 1.170319305214277e-05,
+      "loss": 0.5960506200790405,
+      "step": 584
+    },
+    {
+      "epoch": 2.3346613545816735,
+      "grad_norm": 0.47800812125205994,
+      "learning_rate": 1.1630304085790206e-05,
+      "loss": 0.48045945167541504,
+      "step": 586
+    },
+    {
+      "epoch": 2.3426294820717133,
+      "grad_norm": 0.3794075548648834,
+      "learning_rate": 1.155750725895329e-05,
+      "loss": 0.657084047794342,
+      "step": 588
+    },
+    {
+      "epoch": 2.350597609561753,
+      "grad_norm": 0.29157790541648865,
+      "learning_rate": 1.1484805735984191e-05,
+      "loss": 0.6771900057792664,
+      "step": 590
+    },
+    {
+      "epoch": 2.358565737051793,
+      "grad_norm": 0.378120481967926,
+      "learning_rate": 1.1412202677092388e-05,
+      "loss": 0.7127939462661743,
+      "step": 592
+    },
+    {
+      "epoch": 2.366533864541833,
+      "grad_norm": 0.3749159276485443,
+      "learning_rate": 1.133970123820729e-05,
+      "loss": 0.4554901719093323,
+      "step": 594
+    },
+    {
+      "epoch": 2.3745019920318726,
+      "grad_norm": 0.3079882264137268,
+      "learning_rate": 1.1267304570841067e-05,
+      "loss": 0.6119897961616516,
+      "step": 596
+    },
+    {
+      "epoch": 2.3824701195219125,
+      "grad_norm": 0.2140430063009262,
+      "learning_rate": 1.1195015821951642e-05,
+      "loss": 0.5725736021995544,
+      "step": 598
+    },
+    {
+      "epoch": 2.3904382470119523,
+      "grad_norm": 0.3181394636631012,
+      "learning_rate": 1.1122838133805924e-05,
+      "loss": 0.5081043243408203,
+      "step": 600
+    },
+    {
+      "epoch": 2.398406374501992,
+      "grad_norm": 0.3187297582626343,
+      "learning_rate": 1.1050774643843192e-05,
+      "loss": 0.6516472101211548,
+      "step": 602
+    },
+    {
+      "epoch": 2.406374501992032,
+      "grad_norm": 1.1320936679840088,
+      "learning_rate": 1.0978828484538725e-05,
+      "loss": 0.4262102544307709,
+      "step": 604
+    },
+    {
+      "epoch": 2.414342629482072,
+      "grad_norm": 0.76902174949646,
+      "learning_rate": 1.0907002783267651e-05,
+      "loss": 0.21533596515655518,
+      "step": 606
+    },
+    {
+      "epoch": 2.4223107569721116,
+      "grad_norm": 0.34701016545295715,
+      "learning_rate": 1.083530066216899e-05,
+      "loss": 0.3049370050430298,
+      "step": 608
+    },
+    {
+      "epoch": 2.4302788844621515,
+      "grad_norm": 0.13942857086658478,
+      "learning_rate": 1.0763725238009948e-05,
+      "loss": 0.5161043405532837,
+      "step": 610
+    },
+    {
+      "epoch": 2.4382470119521913,
+      "grad_norm": 0.31895408034324646,
+      "learning_rate": 1.0692279622050424e-05,
+      "loss": 0.5794556140899658,
+      "step": 612
+    },
+    {
+      "epoch": 2.446215139442231,
+      "grad_norm": 0.4094986617565155,
+      "learning_rate": 1.0620966919907802e-05,
+      "loss": 0.3795400857925415,
+      "step": 614
+    },
+    {
+      "epoch": 2.454183266932271,
+      "grad_norm": 0.6053380966186523,
+      "learning_rate": 1.054979023142191e-05,
+      "loss": 0.6777042746543884,
+      "step": 616
+    },
+    {
+      "epoch": 2.462151394422311,
+      "grad_norm": 0.47795283794403076,
+      "learning_rate": 1.0478752650520301e-05,
+      "loss": 0.531360387802124,
+      "step": 618
+    },
+    {
+      "epoch": 2.4701195219123506,
+      "grad_norm": 0.5629068613052368,
+      "learning_rate": 1.0407857265083779e-05,
+      "loss": 0.6984584927558899,
+      "step": 620
+    },
+    {
+      "epoch": 2.4780876494023905,
+      "grad_norm": 0.3025050759315491,
+      "learning_rate": 1.0337107156812135e-05,
+      "loss": 0.6388870477676392,
+      "step": 622
+    },
+    {
+      "epoch": 2.4860557768924303,
+      "grad_norm": 0.3667789399623871,
+      "learning_rate": 1.026650540109022e-05,
+      "loss": 0.6092079281806946,
+      "step": 624
+    },
+    {
+      "epoch": 2.49402390438247,
+      "grad_norm": 1.2361482381820679,
+      "learning_rate": 1.0196055066854265e-05,
+      "loss": 0.3069516718387604,
+      "step": 626
+    },
+    {
+      "epoch": 2.50199203187251,
+      "grad_norm": 0.42347556352615356,
+      "learning_rate": 1.0125759216458467e-05,
+      "loss": 0.6878538727760315,
+      "step": 628
+    },
+    {
+      "epoch": 2.50996015936255,
+      "grad_norm": 0.4100952744483948,
+      "learning_rate": 1.0055620905541865e-05,
+      "loss": 0.23041468858718872,
+      "step": 630
+    },
+    {
+      "epoch": 2.5179282868525896,
+      "grad_norm": 1.4519829750061035,
+      "learning_rate": 9.985643182895546e-06,
+      "loss": 0.3131229877471924,
+      "step": 632
+    },
+    {
+      "epoch": 2.5258964143426295,
+      "grad_norm": 0.4016880989074707,
+      "learning_rate": 9.915829090330097e-06,
+      "loss": 0.6878085136413574,
+      "step": 634
+    },
+    {
+      "epoch": 2.5338645418326693,
+      "grad_norm": 0.30623283982276917,
+      "learning_rate": 9.846181662543385e-06,
+      "loss": 0.6297607421875,
+      "step": 636
+    },
+    {
+      "epoch": 2.541832669322709,
+      "grad_norm": 0.7893882989883423,
+      "learning_rate": 9.77670392698865e-06,
+      "loss": 0.807242214679718,
+      "step": 638
+    },
+    {
+      "epoch": 2.549800796812749,
+      "grad_norm": 0.12514376640319824,
+      "learning_rate": 9.707398903742898e-06,
+      "loss": 0.2793222665786743,
+      "step": 640
+    },
+    {
+      "epoch": 2.557768924302789,
+      "grad_norm": 0.447418749332428,
+      "learning_rate": 9.638269605375646e-06,
+      "loss": 0.6577051877975464,
+      "step": 642
+    },
+    {
+      "epoch": 2.5657370517928286,
+      "grad_norm": 0.25221747159957886,
+      "learning_rate": 9.569319036817942e-06,
+      "loss": 0.6412659883499146,
+      "step": 644
+    },
+    {
+      "epoch": 2.5737051792828685,
+      "grad_norm": 0.29057738184928894,
+      "learning_rate": 9.50055019523175e-06,
+      "loss": 0.5261269211769104,
+      "step": 646
+    },
+    {
+      "epoch": 2.5816733067729083,
+      "grad_norm": 0.38685640692710876,
+      "learning_rate": 9.431966069879694e-06,
+      "loss": 0.655281662940979,
+      "step": 648
+    },
+    {
+      "epoch": 2.589641434262948,
+      "grad_norm": 0.25463399291038513,
+      "learning_rate": 9.363569641995102e-06,
+      "loss": 0.4800514280796051,
+      "step": 650
+    },
+    {
+      "epoch": 2.597609561752988,
+      "grad_norm": 0.599278450012207,
+      "learning_rate": 9.295363884652394e-06,
+      "loss": 0.4531796872615814,
+      "step": 652
+    },
+    {
+      "epoch": 2.605577689243028,
+      "grad_norm": 0.2729853689670563,
+      "learning_rate": 9.227351762637907e-06,
+      "loss": 0.547470211982727,
+      "step": 654
+    },
+    {
+      "epoch": 2.6135458167330676,
+      "grad_norm": 0.37375718355178833,
+      "learning_rate": 9.159536232320955e-06,
+      "loss": 0.23799745738506317,
+      "step": 656
+    },
+    {
+      "epoch": 2.6215139442231075,
+      "grad_norm": 0.3460836112499237,
+      "learning_rate": 9.09192024152537e-06,
+      "loss": 0.4408121705055237,
+      "step": 658
+    },
+    {
+      "epoch": 2.6294820717131473,
+      "grad_norm": 0.2145247757434845,
+      "learning_rate": 9.024506729401333e-06,
+      "loss": 0.608639657497406,
+      "step": 660
+    },
+    {
+      "epoch": 2.637450199203187,
+      "grad_norm": 0.8424778580665588,
+      "learning_rate": 8.957298626297633e-06,
+      "loss": 0.4632888734340668,
+      "step": 662
+    },
+    {
+      "epoch": 2.645418326693227,
+      "grad_norm": 0.42404642701148987,
+      "learning_rate": 8.890298853634292e-06,
+      "loss": 0.42067912220954895,
+      "step": 664
+    },
+    {
+      "epoch": 2.653386454183267,
+      "grad_norm": 0.43092215061187744,
+      "learning_rate": 8.823510323775552e-06,
+      "loss": 0.6638267636299133,
+      "step": 666
+    },
+    {
+      "epoch": 2.6613545816733066,
+      "grad_norm": 0.08587604016065598,
+      "learning_rate": 8.756935939903301e-06,
+      "loss": 0.5093077421188354,
+      "step": 668
+    },
+    {
+      "epoch": 2.6693227091633465,
+      "grad_norm": 0.7821229100227356,
+      "learning_rate": 8.690578595890883e-06,
+      "loss": 0.5566848516464233,
+      "step": 670
+    },
+    {
+      "epoch": 2.6772908366533863,
+      "grad_norm": 0.1694018691778183,
+      "learning_rate": 8.624441176177285e-06,
+      "loss": 0.3764144778251648,
+      "step": 672
+    },
+    {
+      "epoch": 2.685258964143426,
+      "grad_norm": 0.23217065632343292,
+      "learning_rate": 8.558526555641764e-06,
+      "loss": 0.6601039171218872,
+      "step": 674
+    },
+    {
+      "epoch": 2.6932270916334664,
+      "grad_norm": 4.333953380584717,
+      "learning_rate": 8.492837599478875e-06,
+      "loss": 0.472968727350235,
+      "step": 676
+    },
+    {
+      "epoch": 2.7011952191235062,
+      "grad_norm": 0.380391001701355,
+      "learning_rate": 8.427377163073958e-06,
+      "loss": 0.4582363963127136,
+      "step": 678
+    },
+    {
+      "epoch": 2.709163346613546,
+      "grad_norm": 0.3102791905403137,
+      "learning_rate": 8.362148091878962e-06,
+      "loss": 0.7194199562072754,
+      "step": 680
+    },
+    {
+      "epoch": 2.717131474103586,
+      "grad_norm": 0.2712518274784088,
+      "learning_rate": 8.297153221288815e-06,
+      "loss": 0.6685885787010193,
+      "step": 682
+    },
+    {
+      "epoch": 2.7250996015936257,
+      "grad_norm": 0.35200121998786926,
+      "learning_rate": 8.232395376518148e-06,
+      "loss": 0.4935896396636963,
+      "step": 684
+    },
+    {
+      "epoch": 2.7330677290836656,
+      "grad_norm": 8.354119300842285,
+      "learning_rate": 8.167877372478468e-06,
+      "loss": 0.3051784336566925,
+      "step": 686
+    },
+    {
+      "epoch": 2.7410358565737054,
+      "grad_norm": 0.3017679452896118,
+      "learning_rate": 8.103602013655842e-06,
+      "loss": 0.6676462292671204,
+      "step": 688
+    },
+    {
+      "epoch": 2.7490039840637452,
+      "grad_norm": 0.32495346665382385,
+      "learning_rate": 8.039572093988961e-06,
+      "loss": 0.3756825923919678,
+      "step": 690
+    },
+    {
+      "epoch": 2.756972111553785,
+      "grad_norm": 0.3361388146877289,
+      "learning_rate": 7.97579039674771e-06,
+      "loss": 0.7831878662109375,
+      "step": 692
+    },
+    {
+      "epoch": 2.764940239043825,
+      "grad_norm": 0.8197224140167236,
+      "learning_rate": 7.912259694412155e-06,
+      "loss": 0.5665961503982544,
+      "step": 694
+    },
+    {
+      "epoch": 2.7729083665338647,
+      "grad_norm": 0.718382716178894,
+      "learning_rate": 7.848982748552058e-06,
+      "loss": 0.5898396372795105,
+      "step": 696
+    },
+    {
+      "epoch": 2.7808764940239046,
+      "grad_norm": 0.286214679479599,
+      "learning_rate": 7.785962309706834e-06,
+      "loss": 0.6114211082458496,
+      "step": 698
+    },
+    {
+      "epoch": 2.7888446215139444,
+      "grad_norm": 0.2647058665752411,
+      "learning_rate": 7.723201117265967e-06,
+      "loss": 0.7290526032447815,
+      "step": 700
+    },
+    {
+      "epoch": 2.7968127490039842,
+      "grad_norm": 0.17934134602546692,
+      "learning_rate": 7.660701899349961e-06,
+      "loss": 0.41632407903671265,
+      "step": 702
+    },
+    {
+      "epoch": 2.804780876494024,
+      "grad_norm": 0.7281224131584167,
+      "learning_rate": 7.598467372691729e-06,
+      "loss": 0.7692102789878845,
+      "step": 704
+    },
+    {
+      "epoch": 2.812749003984064,
+      "grad_norm": 0.2112652212381363,
+      "learning_rate": 7.536500242518521e-06,
+      "loss": 0.6137795448303223,
+      "step": 706
+    },
+    {
+      "epoch": 2.8207171314741037,
+      "grad_norm": 0.525693416595459,
+      "learning_rate": 7.474803202434319e-06,
+      "loss": 0.8058525323867798,
+      "step": 708
+    },
+    {
+      "epoch": 2.8286852589641436,
+      "grad_norm": 0.2055388242006302,
+      "learning_rate": 7.4133789343027584e-06,
+      "loss": 0.38117560744285583,
+      "step": 710
+    },
+    {
+      "epoch": 2.8366533864541834,
+      "grad_norm": 0.33113518357276917,
+      "learning_rate": 7.352230108130562e-06,
+      "loss": 0.735668957233429,
+      "step": 712
+    },
+    {
+      "epoch": 2.8446215139442232,
+      "grad_norm": 0.28394532203674316,
+      "learning_rate": 7.291359381951435e-06,
+      "loss": 0.4625368118286133,
+      "step": 714
+    },
+    {
+      "epoch": 2.852589641434263,
+      "grad_norm": 0.32944923639297485,
+      "learning_rate": 7.230769401710585e-06,
+      "loss": 0.6521753668785095,
+      "step": 716
+    },
+    {
+      "epoch": 2.860557768924303,
+      "grad_norm": 0.7421508431434631,
+      "learning_rate": 7.1704628011496684e-06,
+      "loss": 0.5228698253631592,
+      "step": 718
+    },
+    {
+      "epoch": 2.8685258964143427,
+      "grad_norm": 0.24052925407886505,
+      "learning_rate": 7.110442201692327e-06,
+      "loss": 0.42685985565185547,
+      "step": 720
+    },
+    {
+      "epoch": 2.8764940239043826,
+      "grad_norm": 0.6366944909095764,
+      "learning_rate": 7.050710212330217e-06,
+      "loss": 0.40909215807914734,
+      "step": 722
+    },
+    {
+      "epoch": 2.8844621513944224,
+      "grad_norm": 0.2380335032939911,
+      "learning_rate": 6.991269429509608e-06,
+      "loss": 0.47712504863739014,
+      "step": 724
+    },
+    {
+      "epoch": 2.8924302788844622,
+      "grad_norm": 0.46672236919403076,
+      "learning_rate": 6.932122437018548e-06,
+      "loss": 0.652793288230896,
+      "step": 726
+    },
+    {
+      "epoch": 2.900398406374502,
+      "grad_norm": 0.2884184718132019,
+      "learning_rate": 6.873271805874504e-06,
+      "loss": 0.39024391770362854,
+      "step": 728
+    },
+    {
+      "epoch": 2.908366533864542,
+      "grad_norm": 0.25529900193214417,
+      "learning_rate": 6.814720094212643e-06,
+      "loss": 0.7011076211929321,
+      "step": 730
+    },
+    {
+      "epoch": 2.9163346613545817,
+      "grad_norm": 0.2143237292766571,
+      "learning_rate": 6.756469847174606e-06,
+      "loss": 0.2097426950931549,
+      "step": 732
+    },
+    {
+      "epoch": 2.9243027888446216,
+      "grad_norm": 0.32660678029060364,
+      "learning_rate": 6.698523596797904e-06,
+      "loss": 0.7196627855300903,
+      "step": 734
+    },
+    {
+      "epoch": 2.9322709163346614,
+      "grad_norm": 0.6449710726737976,
+      "learning_rate": 6.640883861905825e-06,
+      "loss": 0.37205058336257935,
+      "step": 736
+    },
+    {
+      "epoch": 2.9402390438247012,
+      "grad_norm": 0.7962883710861206,
+      "learning_rate": 6.583553147997962e-06,
+      "loss": 0.4908309280872345,
+      "step": 738
+    },
+    {
+      "epoch": 2.948207171314741,
+      "grad_norm": 0.24611425399780273,
+      "learning_rate": 6.526533947141311e-06,
+      "loss": 0.6733062863349915,
+      "step": 740
+    },
+    {
+      "epoch": 2.956175298804781,
+      "grad_norm": 0.8728877902030945,
+      "learning_rate": 6.469828737861922e-06,
+      "loss": 0.281482994556427,
+      "step": 742
+    },
+    {
+      "epoch": 2.9641434262948207,
+      "grad_norm": 0.38738852739334106,
+      "learning_rate": 6.413439985037173e-06,
+      "loss": 0.5737511515617371,
+      "step": 744
+    },
+    {
+      "epoch": 2.9721115537848606,
+      "grad_norm": 0.28229281306266785,
+      "learning_rate": 6.357370139788635e-06,
+      "loss": 0.6209017038345337,
+      "step": 746
+    },
+    {
+      "epoch": 2.9800796812749004,
+      "grad_norm": 0.24411159753799438,
+      "learning_rate": 6.301621639375521e-06,
+      "loss": 0.406149297952652,
+      "step": 748
+    },
+    {
+      "epoch": 2.9880478087649402,
+      "grad_norm": 0.2958669662475586,
+      "learning_rate": 6.246196907088729e-06,
+      "loss": 0.12414422631263733,
+      "step": 750
+    },
+    {
+      "epoch": 2.99601593625498,
+      "grad_norm": 0.23955702781677246,
+      "learning_rate": 6.19109835214552e-06,
+      "loss": 0.6650469899177551,
+      "step": 752
+    },
+    {
+      "epoch": 3.00398406374502,
+      "grad_norm": 0.18620793521404266,
+      "learning_rate": 6.136328369584794e-06,
+      "loss": 0.3096679449081421,
+      "step": 754
+    },
+    {
+      "epoch": 3.0119521912350598,
+      "grad_norm": 0.4313194155693054,
+      "learning_rate": 6.081889340162973e-06,
+      "loss": 0.25961756706237793,
+      "step": 756
+    },
+    {
+      "epoch": 3.0199203187250996,
+      "grad_norm": 0.2690771520137787,
+      "learning_rate": 6.027783630250527e-06,
+      "loss": 0.12224570661783218,
+      "step": 758
+    },
+    {
+      "epoch": 3.0278884462151394,
+      "grad_norm": 0.7186622619628906,
+      "learning_rate": 5.974013591729091e-06,
+      "loss": 0.3025573790073395,
+      "step": 760
+    },
+    {
+      "epoch": 3.0358565737051793,
+      "grad_norm": 1.3814857006072998,
+      "learning_rate": 5.920581561889251e-06,
+      "loss": 0.23363067209720612,
+      "step": 762
+    },
+    {
+      "epoch": 3.043824701195219,
+      "grad_norm": 0.3901160657405853,
+      "learning_rate": 5.867489863328939e-06,
+      "loss": 0.3046700358390808,
+      "step": 764
+    },
+    {
+      "epoch": 3.051792828685259,
+      "grad_norm": 0.37714093923568726,
+      "learning_rate": 5.814740803852481e-06,
+      "loss": 0.16742630302906036,
+      "step": 766
+    },
+    {
+      "epoch": 3.0597609561752988,
+      "grad_norm": 0.46958741545677185,
+      "learning_rate": 5.7623366763702715e-06,
+      "loss": 0.237832710146904,
+      "step": 768
+    },
+    {
+      "epoch": 3.0677290836653386,
+      "grad_norm": 0.4390011727809906,
+      "learning_rate": 5.7102797587991026e-06,
+      "loss": 0.22830596566200256,
+      "step": 770
+    },
+    {
+      "epoch": 3.0756972111553784,
+      "grad_norm": 6.966987133026123,
+      "learning_rate": 5.65857231396315e-06,
+      "loss": 0.35515689849853516,
+      "step": 772
+    },
+    {
+      "epoch": 3.0836653386454183,
+      "grad_norm": 0.658822238445282,
+      "learning_rate": 5.607216589495629e-06,
+      "loss": 0.3950346112251282,
+      "step": 774
+    },
+    {
+      "epoch": 3.091633466135458,
+      "grad_norm": 0.07788458466529846,
+      "learning_rate": 5.556214817741059e-06,
+      "loss": 0.2262239009141922,
+      "step": 776
+    },
+    {
+      "epoch": 3.099601593625498,
+      "grad_norm": 0.30647003650665283,
+      "learning_rate": 5.505569215658268e-06,
+      "loss": 0.21924108266830444,
+      "step": 778
+    },
+    {
+      "epoch": 3.1075697211155378,
+      "grad_norm": 0.3421042859554291,
+      "learning_rate": 5.4552819847239844e-06,
+      "loss": 0.3232831060886383,
+      "step": 780
+    },
+    {
+      "epoch": 3.1155378486055776,
+      "grad_norm": 0.0918380469083786,
+      "learning_rate": 5.4053553108371776e-06,
+      "loss": 0.11867938935756683,
+      "step": 782
+    },
+    {
+      "epoch": 3.1235059760956174,
+      "grad_norm": 0.2789020836353302,
+      "learning_rate": 5.3557913642240165e-06,
+      "loss": 0.3726314604282379,
+      "step": 784
+    },
+    {
+      "epoch": 3.1314741035856573,
+      "grad_norm": 0.3296692669391632,
+      "learning_rate": 5.306592299343544e-06,
+      "loss": 0.23498745262622833,
+      "step": 786
+    },
+    {
+      "epoch": 3.139442231075697,
+      "grad_norm": 0.4843685030937195,
+      "learning_rate": 5.257760254794037e-06,
+      "loss": 0.22427764534950256,
+      "step": 788
+    },
+    {
+      "epoch": 3.147410358565737,
+      "grad_norm": 0.22648558020591736,
+      "learning_rate": 5.20929735322001e-06,
+      "loss": 0.20027917623519897,
+      "step": 790
+    },
+    {
+      "epoch": 3.1553784860557768,
+      "grad_norm": 0.5183139443397522,
+      "learning_rate": 5.1612057012199915e-06,
+      "loss": 0.23067830502986908,
+      "step": 792
+    },
+    {
+      "epoch": 3.1633466135458166,
+      "grad_norm": 0.1671808362007141,
+      "learning_rate": 5.113487389254923e-06,
+      "loss": 0.11641529947519302,
+      "step": 794
+    },
+    {
+      "epoch": 3.1713147410358564,
+      "grad_norm": 0.7252765893936157,
+      "learning_rate": 5.0661444915573055e-06,
+      "loss": 0.26584768295288086,
+      "step": 796
+    },
+    {
+      "epoch": 3.1792828685258963,
+      "grad_norm": 1.4055886268615723,
+      "learning_rate": 5.01917906604103e-06,
+      "loss": 0.2983208894729614,
+      "step": 798
+    },
+    {
+      "epoch": 3.187250996015936,
+      "grad_norm": 0.4246300160884857,
+      "learning_rate": 4.972593154211916e-06,
+      "loss": 0.3397647738456726,
+      "step": 800
+    },
+    {
+      "epoch": 3.195219123505976,
+      "grad_norm": 0.4503802955150604,
+      "learning_rate": 4.9263887810789916e-06,
+      "loss": 0.2282993644475937,
+      "step": 802
+    },
+    {
+      "epoch": 3.2031872509960158,
+      "grad_norm": 0.9323911666870117,
+      "learning_rate": 4.880567955066451e-06,
+      "loss": 0.16186261177062988,
+      "step": 804
+    },
+    {
+      "epoch": 3.2111553784860556,
+      "grad_norm": 0.6153744459152222,
+      "learning_rate": 4.835132667926362e-06,
+      "loss": 0.2332218438386917,
+      "step": 806
+    },
+    {
+      "epoch": 3.2191235059760954,
+      "grad_norm": 0.3493005633354187,
+      "learning_rate": 4.790084894652076e-06,
+      "loss": 0.05802680924534798,
+      "step": 808
+    },
+    {
+      "epoch": 3.2270916334661353,
+      "grad_norm": 0.39290812611579895,
+      "learning_rate": 4.7454265933924004e-06,
+      "loss": 0.381905734539032,
+      "step": 810
+    },
+    {
+      "epoch": 3.235059760956175,
+      "grad_norm": 0.6026656627655029,
+      "learning_rate": 4.701159705366455e-06,
+      "loss": 0.30244818329811096,
+      "step": 812
+    },
+    {
+      "epoch": 3.243027888446215,
+      "grad_norm": 0.3029796779155731,
+      "learning_rate": 4.657286154779308e-06,
+      "loss": 0.25009334087371826,
+      "step": 814
+    },
+    {
+      "epoch": 3.2509960159362548,
+      "grad_norm": 0.5528416633605957,
+      "learning_rate": 4.613807848738337e-06,
+      "loss": 0.0697302594780922,
+      "step": 816
+    },
+    {
+      "epoch": 3.2589641434262946,
+      "grad_norm": 0.3179832696914673,
+      "learning_rate": 4.57072667717031e-06,
+      "loss": 0.3150222599506378,
+      "step": 818
+    },
+    {
+      "epoch": 3.2669322709163344,
+      "grad_norm": 1.380695104598999,
+      "learning_rate": 4.528044512739247e-06,
+      "loss": 0.19252577424049377,
+      "step": 820
+    },
+    {
+      "epoch": 3.2749003984063743,
+      "grad_norm": 0.5292989015579224,
+      "learning_rate": 4.485763210765033e-06,
+      "loss": 0.165231391787529,
+      "step": 822
+    },
+    {
+      "epoch": 3.2828685258964145,
+      "grad_norm": 0.5927901864051819,
+      "learning_rate": 4.443884609142735e-06,
+      "loss": 0.2680898606777191,
+      "step": 824
+    },
+    {
+      "epoch": 3.2908366533864544,
+      "grad_norm": 0.30026665329933167,
+      "learning_rate": 4.402410528262748e-06,
+      "loss": 0.31436988711357117,
+      "step": 826
+    },
+    {
+      "epoch": 3.298804780876494,
+      "grad_norm": 0.30230429768562317,
+      "learning_rate": 4.361342770931641e-06,
+      "loss": 0.40432700514793396,
+      "step": 828
+    },
+    {
+      "epoch": 3.306772908366534,
+      "grad_norm": 0.506352424621582,
+      "learning_rate": 4.320683122293806e-06,
+      "loss": 0.23675966262817383,
+      "step": 830
+    },
+    {
+      "epoch": 3.314741035856574,
+      "grad_norm": 0.34960469603538513,
+      "learning_rate": 4.280433349753851e-06,
+      "loss": 0.2234186828136444,
+      "step": 832
+    },
+    {
+      "epoch": 3.3227091633466137,
+      "grad_norm": 1.093450665473938,
+      "learning_rate": 4.240595202899784e-06,
+      "loss": 0.2236480414867401,
+      "step": 834
+    },
+    {
+      "epoch": 3.3306772908366535,
+      "grad_norm": 1.7021102905273438,
+      "learning_rate": 4.2011704134269536e-06,
+      "loss": 0.244973286986351,
+      "step": 836
+    },
+    {
+      "epoch": 3.3386454183266934,
+      "grad_norm": 1.0127544403076172,
+      "learning_rate": 4.162160695062774e-06,
+      "loss": 0.3762524127960205,
+      "step": 838
+    },
+    {
+      "epoch": 3.346613545816733,
+      "grad_norm": 0.48283928632736206,
+      "learning_rate": 4.123567743492243e-06,
+      "loss": 0.09160395711660385,
+      "step": 840
+    },
+    {
+      "epoch": 3.354581673306773,
+      "grad_norm": 0.3954333961009979,
+      "learning_rate": 4.085393236284226e-06,
+      "loss": 0.20918384194374084,
+      "step": 842
+    },
+    {
+      "epoch": 3.362549800796813,
+      "grad_norm": 0.3748583197593689,
+      "learning_rate": 4.0476388328185306e-06,
+      "loss": 0.24666330218315125,
+      "step": 844
+    },
+    {
+      "epoch": 3.3705179282868527,
+      "grad_norm": 0.37042000889778137,
+      "learning_rate": 4.010306174213786e-06,
+      "loss": 0.33929553627967834,
+      "step": 846
+    },
+    {
+      "epoch": 3.3784860557768925,
+      "grad_norm": 0.38942667841911316,
+      "learning_rate": 3.973396883256098e-06,
+      "loss": 0.3574102222919464,
+      "step": 848
+    },
+    {
+      "epoch": 3.3864541832669324,
+      "grad_norm": 0.4091942310333252,
+      "learning_rate": 3.9369125643285164e-06,
+      "loss": 0.23076759278774261,
+      "step": 850
+    },
+    {
+      "epoch": 3.394422310756972,
+      "grad_norm": 0.31176552176475525,
+      "learning_rate": 3.900854803341286e-06,
+      "loss": 0.3034666180610657,
+      "step": 852
+    },
+    {
+      "epoch": 3.402390438247012,
+      "grad_norm": 0.341012567281723,
+      "learning_rate": 3.865225167662926e-06,
+      "loss": 0.3238814175128937,
+      "step": 854
+    },
+    {
+      "epoch": 3.410358565737052,
+      "grad_norm": 0.9658647179603577,
+      "learning_rate": 3.830025206052076e-06,
+      "loss": 0.29980388283729553,
+      "step": 856
+    },
+    {
+      "epoch": 3.4183266932270917,
+      "grad_norm": 0.28022247552871704,
+      "learning_rate": 3.7952564485901993e-06,
+      "loss": 0.3209022283554077,
+      "step": 858
+    },
+    {
+      "epoch": 3.4262948207171315,
+      "grad_norm": 0.30447283387184143,
+      "learning_rate": 3.760920406615049e-06,
+      "loss": 0.31163108348846436,
+      "step": 860
+    },
+    {
+      "epoch": 3.4342629482071714,
+      "grad_norm": 0.460216760635376,
+      "learning_rate": 3.727018572654993e-06,
+      "loss": 0.19528982043266296,
+      "step": 862
+    },
+    {
+      "epoch": 3.442231075697211,
+      "grad_norm": 0.5468026995658875,
+      "learning_rate": 3.6935524203641296e-06,
+      "loss": 0.20554818212985992,
+      "step": 864
+    },
+    {
+      "epoch": 3.450199203187251,
+      "grad_norm": 0.3372550308704376,
+      "learning_rate": 3.660523404458213e-06,
+      "loss": 0.35659119486808777,
+      "step": 866
+    },
+    {
+      "epoch": 3.458167330677291,
+      "grad_norm": 0.2969035804271698,
+      "learning_rate": 3.6279329606514515e-06,
+      "loss": 0.2657085657119751,
+      "step": 868
+    },
+    {
+      "epoch": 3.4661354581673307,
+      "grad_norm": 0.3335762619972229,
+      "learning_rate": 3.5957825055940745e-06,
+      "loss": 0.25995686650276184,
+      "step": 870
+    },
+    {
+      "epoch": 3.4741035856573705,
+      "grad_norm": 0.07810402661561966,
+      "learning_rate": 3.5640734368107665e-06,
+      "loss": 0.21163897216320038,
+      "step": 872
+    },
+    {
+      "epoch": 3.4820717131474104,
+      "grad_norm": 0.06156105175614357,
+      "learning_rate": 3.532807132639909e-06,
+      "loss": 0.1695207953453064,
+      "step": 874
+    },
+    {
+      "epoch": 3.49003984063745,
+      "grad_norm": 0.29371005296707153,
+      "learning_rate": 3.5019849521736703e-06,
+      "loss": 0.1652969866991043,
+      "step": 876
+    },
+    {
+      "epoch": 3.49800796812749,
+      "grad_norm": 0.29471254348754883,
+      "learning_rate": 3.4716082351989357e-06,
+      "loss": 0.2616461217403412,
+      "step": 878
+    },
+    {
+      "epoch": 3.50597609561753,
+      "grad_norm": 0.10254102945327759,
+      "learning_rate": 3.4416783021390577e-06,
+      "loss": 0.1497431844472885,
+      "step": 880
+    },
+    {
+      "epoch": 3.5139442231075697,
+      "grad_norm": 0.3706572949886322,
+      "learning_rate": 3.412196453996469e-06,
+      "loss": 0.10484407097101212,
+      "step": 882
+    },
+    {
+      "epoch": 3.5219123505976095,
+      "grad_norm": 0.28233277797698975,
+      "learning_rate": 3.3831639722961196e-06,
+      "loss": 0.04023301973938942,
+      "step": 884
+    },
+    {
+      "epoch": 3.5298804780876494,
+      "grad_norm": 0.5272843837738037,
+      "learning_rate": 3.354582119029784e-06,
+      "loss": 0.3029521703720093,
+      "step": 886
+    },
+    {
+      "epoch": 3.537848605577689,
+      "grad_norm": 0.4497409462928772,
+      "learning_rate": 3.326452136601192e-06,
+      "loss": 0.35895416140556335,
+      "step": 888
+    },
+    {
+      "epoch": 3.545816733067729,
+      "grad_norm": 0.3548281788825989,
+      "learning_rate": 3.298775247772035e-06,
+      "loss": 0.17913445830345154,
+      "step": 890
+    },
+    {
+      "epoch": 3.553784860557769,
+      "grad_norm": 0.6090046167373657,
+      "learning_rate": 3.27155265560881e-06,
+      "loss": 0.20643927156925201,
+      "step": 892
+    },
+    {
+      "epoch": 3.5617529880478087,
+      "grad_norm": 0.2786542773246765,
+      "learning_rate": 3.2447855434305134e-06,
+      "loss": 0.5030993223190308,
+      "step": 894
+    },
+    {
+      "epoch": 3.5697211155378485,
+      "grad_norm": 0.14362268149852753,
+      "learning_rate": 3.218475074757224e-06,
+      "loss": 0.06106302887201309,
+      "step": 896
+    },
+    {
+      "epoch": 3.5776892430278884,
+      "grad_norm": 0.259156197309494,
+      "learning_rate": 3.192622393259518e-06,
+      "loss": 0.16078512370586395,
+      "step": 898
+    },
+    {
+      "epoch": 3.585657370517928,
+      "grad_norm": 0.3713056743144989,
+      "learning_rate": 3.1672286227087457e-06,
+      "loss": 0.08721793442964554,
+      "step": 900
+    },
+    {
+      "epoch": 3.593625498007968,
+      "grad_norm": 0.470853328704834,
+      "learning_rate": 3.1422948669282013e-06,
+      "loss": 0.06528808921575546,
+      "step": 902
+    },
+    {
+      "epoch": 3.601593625498008,
+      "grad_norm": 0.4089232087135315,
+      "learning_rate": 3.1178222097451264e-06,
+      "loss": 0.2283235788345337,
+      "step": 904
+    },
+    {
+      "epoch": 3.6095617529880477,
+      "grad_norm": 0.33979326486587524,
+      "learning_rate": 3.09381171494361e-06,
+      "loss": 0.15290184319019318,
+      "step": 906
+    },
+    {
+      "epoch": 3.6175298804780875,
+      "grad_norm": 0.31778720021247864,
+      "learning_rate": 3.0702644262183325e-06,
+      "loss": 0.3788220286369324,
+      "step": 908
+    },
+    {
+      "epoch": 3.6254980079681274,
+      "grad_norm": 0.484840989112854,
+      "learning_rate": 3.047181367129219e-06,
+      "loss": 0.2795168459415436,
+      "step": 910
+    },
+    {
+      "epoch": 3.633466135458167,
+      "grad_norm": 0.08123687654733658,
+      "learning_rate": 3.0245635410569265e-06,
+      "loss": 0.038957491517066956,
+      "step": 912
+    },
+    {
+      "epoch": 3.641434262948207,
+      "grad_norm": 0.3782183825969696,
+      "learning_rate": 3.002411931159237e-06,
+      "loss": 0.2506720721721649,
+      "step": 914
+    },
+    {
+      "epoch": 3.649402390438247,
+      "grad_norm": 0.35183200240135193,
+      "learning_rate": 2.980727500328328e-06,
+      "loss": 0.2860797345638275,
+      "step": 916
+    },
+    {
+      "epoch": 3.6573705179282867,
+      "grad_norm": 0.5603599548339844,
+      "learning_rate": 2.9595111911489044e-06,
+      "loss": 0.05544031411409378,
+      "step": 918
+    },
+    {
+      "epoch": 3.6653386454183265,
+      "grad_norm": 0.34753990173339844,
+      "learning_rate": 2.9387639258572393e-06,
+      "loss": 0.2869701683521271,
+      "step": 920
+    },
+    {
+      "epoch": 3.6733067729083664,
+      "grad_norm": 0.16297142207622528,
+      "learning_rate": 2.9184866063010715e-06,
+      "loss": 0.0460997000336647,
+      "step": 922
+    },
+    {
+      "epoch": 3.681274900398406,
+      "grad_norm": 0.8638817071914673,
+      "learning_rate": 2.898680113900416e-06,
+      "loss": 0.2628190815448761,
+      "step": 924
+    },
+    {
+      "epoch": 3.6892430278884465,
+      "grad_norm": 0.35263216495513916,
+      "learning_rate": 2.8793453096092467e-06,
+      "loss": 0.08740659058094025,
+      "step": 926
+    },
+    {
+      "epoch": 3.6972111553784863,
+      "grad_norm": 0.2342199981212616,
+      "learning_rate": 2.860483033878065e-06,
+      "loss": 0.1996232569217682,
+      "step": 928
+    },
+    {
+      "epoch": 3.705179282868526,
+      "grad_norm": 0.08421018719673157,
+      "learning_rate": 2.8420941066173814e-06,
+      "loss": 0.20934821665287018,
+      "step": 930
+    },
+    {
+      "epoch": 3.713147410358566,
+      "grad_norm": 0.343291699886322,
+      "learning_rate": 2.8241793271620595e-06,
+      "loss": 0.30004775524139404,
+      "step": 932
+    },
+    {
+      "epoch": 3.721115537848606,
+      "grad_norm": 0.40662598609924316,
+      "learning_rate": 2.806739474236583e-06,
+      "loss": 0.17531633377075195,
+      "step": 934
+    },
+    {
+      "epoch": 3.7290836653386457,
+      "grad_norm": 0.41496917605400085,
+      "learning_rate": 2.7897753059211975e-06,
+      "loss": 0.3278602063655853,
+      "step": 936
+    },
+    {
+      "epoch": 3.7370517928286855,
+      "grad_norm": 0.559548556804657,
+      "learning_rate": 2.7732875596189618e-06,
+      "loss": 0.11036017537117004,
+      "step": 938
+    },
+    {
+      "epoch": 3.7450199203187253,
+      "grad_norm": 0.31999555230140686,
+      "learning_rate": 2.7572769520236975e-06,
+      "loss": 0.34991517663002014,
+      "step": 940
+    },
+    {
+      "epoch": 3.752988047808765,
+      "grad_norm": 0.23487690091133118,
+      "learning_rate": 2.7417441790888235e-06,
+      "loss": 0.17355544865131378,
+      "step": 942
+    },
+    {
+      "epoch": 3.760956175298805,
+      "grad_norm": 0.1940401941537857,
+      "learning_rate": 2.7266899159971187e-06,
+      "loss": 0.1706204116344452,
+      "step": 944
+    },
+    {
+      "epoch": 3.768924302788845,
+      "grad_norm": 0.3785277009010315,
+      "learning_rate": 2.7121148171313677e-06,
+      "loss": 0.39207062125205994,
+      "step": 946
+    },
+    {
+      "epoch": 3.7768924302788847,
+      "grad_norm": 0.343730628490448,
+      "learning_rate": 2.698019516045913e-06,
+      "loss": 0.2191389501094818,
+      "step": 948
+    },
+    {
+      "epoch": 3.7848605577689245,
+      "grad_norm": 0.2559804916381836,
+      "learning_rate": 2.6844046254391136e-06,
+      "loss": 0.36427849531173706,
+      "step": 950
+    },
+    {
+      "epoch": 3.7928286852589643,
+      "grad_norm": 0.25515085458755493,
+      "learning_rate": 2.671270737126719e-06,
+      "loss": 0.30214783549308777,
+      "step": 952
+    },
+    {
+      "epoch": 3.800796812749004,
+      "grad_norm": 0.325765460729599,
+      "learning_rate": 2.6586184220161427e-06,
+      "loss": 0.18529954552650452,
+      "step": 954
+    },
+    {
+      "epoch": 3.808764940239044,
+      "grad_norm": 0.6795303821563721,
+      "learning_rate": 2.6464482300816413e-06,
+      "loss": 0.37121716141700745,
+      "step": 956
+    },
+    {
+      "epoch": 3.816733067729084,
+      "grad_norm": 0.1896744966506958,
+      "learning_rate": 2.6347606903404116e-06,
+      "loss": 0.1599966585636139,
+      "step": 958
+    },
+    {
+      "epoch": 3.8247011952191237,
+      "grad_norm": 0.359385222196579,
+      "learning_rate": 2.623556310829593e-06,
+      "loss": 0.11625859141349792,
+      "step": 960
+    },
+    {
+      "epoch": 3.8326693227091635,
+      "grad_norm": 0.09733131527900696,
+      "learning_rate": 2.6128355785841883e-06,
+      "loss": 0.2088605761528015,
+      "step": 962
+    },
+    {
+      "epoch": 3.8406374501992033,
+      "grad_norm": 0.3594829738140106,
+      "learning_rate": 2.6025989596158866e-06,
+      "loss": 0.41164886951446533,
+      "step": 964
+    },
+    {
+      "epoch": 3.848605577689243,
+      "grad_norm": 0.5380414724349976,
+      "learning_rate": 2.5928468988928117e-06,
+      "loss": 0.17345690727233887,
+      "step": 966
+    },
+    {
+      "epoch": 3.856573705179283,
+      "grad_norm": 0.5892635583877563,
+      "learning_rate": 2.5835798203201787e-06,
+      "loss": 0.2570192217826843,
+      "step": 968
+    },
+    {
+      "epoch": 3.864541832669323,
+      "grad_norm": 0.3574495315551758,
+      "learning_rate": 2.574798126721867e-06,
+      "loss": 0.23841306567192078,
+      "step": 970
+    },
+    {
+      "epoch": 3.8725099601593627,
+      "grad_norm": 0.05246090516448021,
+      "learning_rate": 2.5665021998229085e-06,
+      "loss": 0.22528496384620667,
+      "step": 972
+    },
+    {
+      "epoch": 3.8804780876494025,
+      "grad_norm": 1.4794021844863892,
+      "learning_rate": 2.5586924002329033e-06,
+      "loss": 0.20363985002040863,
+      "step": 974
+    },
+    {
+      "epoch": 3.8884462151394423,
+      "grad_norm": 0.36268168687820435,
+      "learning_rate": 2.551369067430328e-06,
+      "loss": 0.3567611277103424,
+      "step": 976
+    },
+    {
+      "epoch": 3.896414342629482,
+      "grad_norm": 0.6210702657699585,
+      "learning_rate": 2.544532519747799e-06,
+      "loss": 0.24332654476165771,
+      "step": 978
+    },
+    {
+      "epoch": 3.904382470119522,
+      "grad_norm": 0.08069682866334915,
+      "learning_rate": 2.5381830543582185e-06,
+      "loss": 0.16302455961704254,
+      "step": 980
+    },
+    {
+      "epoch": 3.912350597609562,
+      "grad_norm": 0.5772701501846313,
+      "learning_rate": 2.5323209472618684e-06,
+      "loss": 0.22316807508468628,
+      "step": 982
+    },
+    {
+      "epoch": 3.9203187250996017,
+      "grad_norm": 0.4512209892272949,
+      "learning_rate": 2.526946453274405e-06,
+      "loss": 0.3951210379600525,
+      "step": 984
+    },
+    {
+      "epoch": 3.9282868525896415,
+      "grad_norm": 0.37502825260162354,
+      "learning_rate": 2.522059806015788e-06,
+      "loss": 0.10022931545972824,
+      "step": 986
+    },
+    {
+      "epoch": 3.9362549800796813,
+      "grad_norm": 7.970139026641846,
+      "learning_rate": 2.5176612179001236e-06,
+      "loss": 0.3307144045829773,
+      "step": 988
+    },
+    {
+      "epoch": 3.944223107569721,
+      "grad_norm": 0.0624762661755085,
+      "learning_rate": 2.5137508801264286e-06,
+      "loss": 0.19763189554214478,
+      "step": 990
+    },
+    {
+      "epoch": 3.952191235059761,
+      "grad_norm": 0.2932847738265991,
+      "learning_rate": 2.5103289626703247e-06,
+      "loss": 0.30540770292282104,
+      "step": 992
+    },
+    {
+      "epoch": 3.960159362549801,
+      "grad_norm": 0.2334122210741043,
+      "learning_rate": 2.5073956142766468e-06,
+      "loss": 0.11537239700555801,
+      "step": 994
+    },
+    {
+      "epoch": 3.9681274900398407,
+      "grad_norm": 0.6144185066223145,
+      "learning_rate": 2.504950962452975e-06,
+      "loss": 0.2402108907699585,
+      "step": 996
+    },
+    {
+      "epoch": 3.9760956175298805,
+      "grad_norm": 0.09196891635656357,
+      "learning_rate": 2.5029951134640978e-06,
+      "loss": 0.20369143784046173,
+      "step": 998
+    },
+    {
+      "epoch": 3.9840637450199203,
+      "grad_norm": 0.3909507095813751,
+      "learning_rate": 2.501528152327387e-06,
+      "loss": 0.15496620535850525,
+      "step": 1000
+    },
+    {
+      "epoch": 3.99203187250996,
+      "grad_norm": 0.3289106488227844,
+      "learning_rate": 2.500550142809107e-06,
+      "loss": 0.2950751781463623,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.10831116139888763,
+      "learning_rate": 2.500061127421641e-06,
+      "loss": 0.11359266936779022,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0,
+      "step": 1004,
+      "total_flos": 4.038502240003031e+18,
+      "train_loss": 0.7041097417681578,
+      "train_runtime": 10571.9353,
+      "train_samples_per_second": 5.698,
+      "train_steps_per_second": 0.095
+    }
+  ],
+  "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": 4.038502240003031e+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 sha256:98ca1a3153d0bb381eded7d815a0d6c846f865dfd739454844594d8ea37ee83d
+size 5649