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", 5e-6), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_adamw_lr", 5e-6), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_adamw_lr", 1e-5), merger_lr),
+    }
+- train_batch_size: 1
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- gradient_accumulation_steps: 60
+- 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", 5e-6), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_adamw_lr", 5e-6), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_adamw_lr", 1e-5), merger_lr),
+    }
+- train_batch_size: 1
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- gradient_accumulation_steps: 60
+- 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": 2246.4632334123307,
+    "epoch": 4.0,
+    "total_flos": 3.954132003103703e+18,
+    "train_loss": 0.909145103641359,
+    "train_runtime": 34871.4276,
+    "train_samples_per_second": 1.727,
+    "train_steps_per_second": 0.029
+}

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 sha256:8277a76446c1fbdaf8d91efeb3a84d811ae06e83305f3883699713293ce6f142
+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": 2246.4632334123307,
+    "epoch": 4.0,
+    "total_flos": 3.954132003103703e+18,
+    "train_loss": 0.909145103641359,
+    "train_runtime": 34871.4276,
+    "train_samples_per_second": 1.727,
+    "train_steps_per_second": 0.029
+}

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": 0.36455339193344116,
+      "learning_rate": 9.803921568627452e-08,
+      "loss": 2.1478986740112305,
+      "step": 2
+    },
+    {
+      "epoch": 0.01593625498007968,
+      "grad_norm": 0.20996835827827454,
+      "learning_rate": 2.9411764705882356e-07,
+      "loss": 2.0985677242279053,
+      "step": 4
+    },
+    {
+      "epoch": 0.02390438247011952,
+      "grad_norm": 1.8723046779632568,
+      "learning_rate": 4.901960784313725e-07,
+      "loss": 2.2998197078704834,
+      "step": 6
+    },
+    {
+      "epoch": 0.03187250996015936,
+      "grad_norm": 0.2750701308250427,
+      "learning_rate": 6.862745098039217e-07,
+      "loss": 1.9246314764022827,
+      "step": 8
+    },
+    {
+      "epoch": 0.0398406374501992,
+      "grad_norm": 0.40245550870895386,
+      "learning_rate": 8.823529411764707e-07,
+      "loss": 1.577760100364685,
+      "step": 10
+    },
+    {
+      "epoch": 0.04780876494023904,
+      "grad_norm": 0.39444074034690857,
+      "learning_rate": 1.0784313725490197e-06,
+      "loss": 1.8360772132873535,
+      "step": 12
+    },
+    {
+      "epoch": 0.055776892430278883,
+      "grad_norm": 0.24191120266914368,
+      "learning_rate": 1.2745098039215686e-06,
+      "loss": 2.811206340789795,
+      "step": 14
+    },
+    {
+      "epoch": 0.06374501992031872,
+      "grad_norm": 2.497779130935669,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 4.562131881713867,
+      "step": 16
+    },
+    {
+      "epoch": 0.07171314741035857,
+      "grad_norm": 0.3985731899738312,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.8284410238265991,
+      "step": 18
+    },
+    {
+      "epoch": 0.0796812749003984,
+      "grad_norm": 0.22641825675964355,
+      "learning_rate": 1.8627450980392158e-06,
+      "loss": 1.699713110923767,
+      "step": 20
+    },
+    {
+      "epoch": 0.08764940239043825,
+      "grad_norm": 0.2197515368461609,
+      "learning_rate": 2.058823529411765e-06,
+      "loss": 1.698561429977417,
+      "step": 22
+    },
+    {
+      "epoch": 0.09561752988047809,
+      "grad_norm": 0.4532564580440521,
+      "learning_rate": 2.254901960784314e-06,
+      "loss": 1.4896349906921387,
+      "step": 24
+    },
+    {
+      "epoch": 0.10358565737051793,
+      "grad_norm": 0.11141212284564972,
+      "learning_rate": 2.450980392156863e-06,
+      "loss": 1.5941441059112549,
+      "step": 26
+    },
+    {
+      "epoch": 0.11155378486055777,
+      "grad_norm": 0.33281439542770386,
+      "learning_rate": 2.647058823529412e-06,
+      "loss": 1.6722030639648438,
+      "step": 28
+    },
+    {
+      "epoch": 0.11952191235059761,
+      "grad_norm": 0.3046768009662628,
+      "learning_rate": 2.843137254901961e-06,
+      "loss": 1.4514604806900024,
+      "step": 30
+    },
+    {
+      "epoch": 0.12749003984063745,
+      "grad_norm": 0.3599211275577545,
+      "learning_rate": 3.03921568627451e-06,
+      "loss": 1.2411736249923706,
+      "step": 32
+    },
+    {
+      "epoch": 0.13545816733067728,
+      "grad_norm": 0.46319112181663513,
+      "learning_rate": 3.2352941176470594e-06,
+      "loss": 1.3915144205093384,
+      "step": 34
+    },
+    {
+      "epoch": 0.14342629482071714,
+      "grad_norm": 0.16933007538318634,
+      "learning_rate": 3.431372549019608e-06,
+      "loss": 1.5070427656173706,
+      "step": 36
+    },
+    {
+      "epoch": 0.15139442231075698,
+      "grad_norm": 0.2670910656452179,
+      "learning_rate": 3.6274509803921573e-06,
+      "loss": 1.3164114952087402,
+      "step": 38
+    },
+    {
+      "epoch": 0.1593625498007968,
+      "grad_norm": 0.17111174762248993,
+      "learning_rate": 3.8235294117647055e-06,
+      "loss": 1.4080593585968018,
+      "step": 40
+    },
+    {
+      "epoch": 0.16733067729083664,
+      "grad_norm": 1.2425928115844727,
+      "learning_rate": 4.019607843137255e-06,
+      "loss": 0.7937950491905212,
+      "step": 42
+    },
+    {
+      "epoch": 0.1752988047808765,
+      "grad_norm": 0.1409052163362503,
+      "learning_rate": 4.215686274509805e-06,
+      "loss": 1.0431537628173828,
+      "step": 44
+    },
+    {
+      "epoch": 0.18326693227091634,
+      "grad_norm": 0.4378255307674408,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 0.9699471592903137,
+      "step": 46
+    },
+    {
+      "epoch": 0.19123505976095617,
+      "grad_norm": 0.44866812229156494,
+      "learning_rate": 4.607843137254902e-06,
+      "loss": 1.2296644449234009,
+      "step": 48
+    },
+    {
+      "epoch": 0.199203187250996,
+      "grad_norm": 0.16275790333747864,
+      "learning_rate": 4.803921568627452e-06,
+      "loss": 1.437253475189209,
+      "step": 50
+    },
+    {
+      "epoch": 0.20717131474103587,
+      "grad_norm": 0.14053893089294434,
+      "learning_rate": 5e-06,
+      "loss": 1.3753666877746582,
+      "step": 52
+    },
+    {
+      "epoch": 0.2151394422310757,
+      "grad_norm": 1.237989068031311,
+      "learning_rate": 4.999951098195543e-06,
+      "loss": 1.0511358976364136,
+      "step": 54
+    },
+    {
+      "epoch": 0.22310756972111553,
+      "grad_norm": 0.42440542578697205,
+      "learning_rate": 4.9998043949078495e-06,
+      "loss": 1.5662118196487427,
+      "step": 56
+    },
+    {
+      "epoch": 0.23107569721115537,
+      "grad_norm": 0.19803501665592194,
+      "learning_rate": 4.9995598965138565e-06,
+      "loss": 1.395676851272583,
+      "step": 58
+    },
+    {
+      "epoch": 0.23904382470119523,
+      "grad_norm": 0.16477660834789276,
+      "learning_rate": 4.9992176136414875e-06,
+      "loss": 0.8043187856674194,
+      "step": 60
+    },
+    {
+      "epoch": 0.24701195219123506,
+      "grad_norm": 0.1683095246553421,
+      "learning_rate": 4.998777561169186e-06,
+      "loss": 1.400430679321289,
+      "step": 62
+    },
+    {
+      "epoch": 0.2549800796812749,
+      "grad_norm": 1.3216127157211304,
+      "learning_rate": 4.998239758225272e-06,
+      "loss": 1.118573784828186,
+      "step": 64
+    },
+    {
+      "epoch": 0.26294820717131473,
+      "grad_norm": 0.1261695772409439,
+      "learning_rate": 4.99760422818711e-06,
+      "loss": 1.320631742477417,
+      "step": 66
+    },
+    {
+      "epoch": 0.27091633466135456,
+      "grad_norm": 0.22709761559963226,
+      "learning_rate": 4.9968709986800925e-06,
+      "loss": 0.7775652408599854,
+      "step": 68
+    },
+    {
+      "epoch": 0.2788844621513944,
+      "grad_norm": 0.33978021144866943,
+      "learning_rate": 4.996040101576436e-06,
+      "loss": 1.6184344291687012,
+      "step": 70
+    },
+    {
+      "epoch": 0.2868525896414343,
+      "grad_norm": 0.18861767649650574,
+      "learning_rate": 4.995111572993804e-06,
+      "loss": 0.7689253091812134,
+      "step": 72
+    },
+    {
+      "epoch": 0.2948207171314741,
+      "grad_norm": 0.23436765372753143,
+      "learning_rate": 4.994085453293725e-06,
+      "loss": 1.1867567300796509,
+      "step": 74
+    },
+    {
+      "epoch": 0.30278884462151395,
+      "grad_norm": 0.14257368445396423,
+      "learning_rate": 4.992961787079848e-06,
+      "loss": 1.4653486013412476,
+      "step": 76
+    },
+    {
+      "epoch": 0.3107569721115538,
+      "grad_norm": 1.1300361156463623,
+      "learning_rate": 4.991740623196001e-06,
+      "loss": 0.9205513596534729,
+      "step": 78
+    },
+    {
+      "epoch": 0.3187250996015936,
+      "grad_norm": 0.09597816318273544,
+      "learning_rate": 4.990422014724068e-06,
+      "loss": 0.8009955883026123,
+      "step": 80
+    },
+    {
+      "epoch": 0.32669322709163345,
+      "grad_norm": 0.2317924052476883,
+      "learning_rate": 4.989006018981675e-06,
+      "loss": 1.3265870809555054,
+      "step": 82
+    },
+    {
+      "epoch": 0.3346613545816733,
+      "grad_norm": 0.416987806558609,
+      "learning_rate": 4.987492697519712e-06,
+      "loss": 0.8422784209251404,
+      "step": 84
+    },
+    {
+      "epoch": 0.3426294820717131,
+      "grad_norm": 0.2015259563922882,
+      "learning_rate": 4.985882116119643e-06,
+      "loss": 1.3318895101547241,
+      "step": 86
+    },
+    {
+      "epoch": 0.350597609561753,
+      "grad_norm": 0.7740209102630615,
+      "learning_rate": 4.9841743447906576e-06,
+      "loss": 0.8810979723930359,
+      "step": 88
+    },
+    {
+      "epoch": 0.35856573705179284,
+      "grad_norm": 0.17352622747421265,
+      "learning_rate": 4.982369457766621e-06,
+      "loss": 1.3504712581634521,
+      "step": 90
+    },
+    {
+      "epoch": 0.3665338645418327,
+      "grad_norm": 1.4900528192520142,
+      "learning_rate": 4.980467533502853e-06,
+      "loss": 1.1341170072555542,
+      "step": 92
+    },
+    {
+      "epoch": 0.3745019920318725,
+      "grad_norm": 0.15929292142391205,
+      "learning_rate": 4.978468654672709e-06,
+      "loss": 1.2599682807922363,
+      "step": 94
+    },
+    {
+      "epoch": 0.38247011952191234,
+      "grad_norm": 0.21968646347522736,
+      "learning_rate": 4.976372908163999e-06,
+      "loss": 1.139507532119751,
+      "step": 96
+    },
+    {
+      "epoch": 0.3904382470119522,
+      "grad_norm": 0.14162515103816986,
+      "learning_rate": 4.974180385075195e-06,
+      "loss": 1.2863703966140747,
+      "step": 98
+    },
+    {
+      "epoch": 0.398406374501992,
+      "grad_norm": 0.13868273794651031,
+      "learning_rate": 4.971891180711488e-06,
+      "loss": 0.8605481386184692,
+      "step": 100
+    },
+    {
+      "epoch": 0.4063745019920319,
+      "grad_norm": 0.2086641937494278,
+      "learning_rate": 4.9695053945806295e-06,
+      "loss": 1.2967108488082886,
+      "step": 102
+    },
+    {
+      "epoch": 0.41434262948207173,
+      "grad_norm": 0.1979150027036667,
+      "learning_rate": 4.9670231303886194e-06,
+      "loss": 0.7460129261016846,
+      "step": 104
+    },
+    {
+      "epoch": 0.42231075697211157,
+      "grad_norm": 1.2564876079559326,
+      "learning_rate": 4.9644444960351865e-06,
+      "loss": 0.7426438331604004,
+      "step": 106
+    },
+    {
+      "epoch": 0.4302788844621514,
+      "grad_norm": 0.17365410923957825,
+      "learning_rate": 4.9617696036091095e-06,
+      "loss": 1.0151516199111938,
+      "step": 108
+    },
+    {
+      "epoch": 0.43824701195219123,
+      "grad_norm": 0.14322529733181,
+      "learning_rate": 4.958998569383335e-06,
+      "loss": 1.302347183227539,
+      "step": 110
+    },
+    {
+      "epoch": 0.44621513944223107,
+      "grad_norm": 0.3185536861419678,
+      "learning_rate": 4.956131513809928e-06,
+      "loss": 0.8034355044364929,
+      "step": 112
+    },
+    {
+      "epoch": 0.4541832669322709,
+      "grad_norm": 0.22524043917655945,
+      "learning_rate": 4.953168561514838e-06,
+      "loss": 1.1029731035232544,
+      "step": 114
+    },
+    {
+      "epoch": 0.46215139442231074,
+      "grad_norm": 0.17002734541893005,
+      "learning_rate": 4.950109841292475e-06,
+      "loss": 1.3041455745697021,
+      "step": 116
+    },
+    {
+      "epoch": 0.4701195219123506,
+      "grad_norm": 0.1084880530834198,
+      "learning_rate": 4.946955486100119e-06,
+      "loss": 1.370314598083496,
+      "step": 118
+    },
+    {
+      "epoch": 0.47808764940239046,
+      "grad_norm": 0.1851903349161148,
+      "learning_rate": 4.943705633052134e-06,
+      "loss": 1.2433099746704102,
+      "step": 120
+    },
+    {
+      "epoch": 0.4860557768924303,
+      "grad_norm": 0.14510802924633026,
+      "learning_rate": 4.940360423414013e-06,
+      "loss": 0.9467571973800659,
+      "step": 122
+    },
+    {
+      "epoch": 0.4940239043824701,
+      "grad_norm": 0.2504097521305084,
+      "learning_rate": 4.936920002596233e-06,
+      "loss": 1.062861680984497,
+      "step": 124
+    },
+    {
+      "epoch": 0.50199203187251,
+      "grad_norm": 0.23112048208713531,
+      "learning_rate": 4.9333845201479375e-06,
+      "loss": 0.9113219380378723,
+      "step": 126
+    },
+    {
+      "epoch": 0.5099601593625498,
+      "grad_norm": 0.13006103038787842,
+      "learning_rate": 4.9297541297504345e-06,
+      "loss": 1.3093539476394653,
+      "step": 128
+    },
+    {
+      "epoch": 0.5179282868525896,
+      "grad_norm": 0.46530863642692566,
+      "learning_rate": 4.9260289892105164e-06,
+      "loss": 0.9284161925315857,
+      "step": 130
+    },
+    {
+      "epoch": 0.5258964143426295,
+      "grad_norm": 0.12075220793485641,
+      "learning_rate": 4.922209260453599e-06,
+      "loss": 0.7887148261070251,
+      "step": 132
+    },
+    {
+      "epoch": 0.5338645418326693,
+      "grad_norm": 0.15673382580280304,
+      "learning_rate": 4.9182951095166875e-06,
+      "loss": 0.8739302754402161,
+      "step": 134
+    },
+    {
+      "epoch": 0.5418326693227091,
+      "grad_norm": 0.21066538989543915,
+      "learning_rate": 4.914286706541153e-06,
+      "loss": 1.2850011587142944,
+      "step": 136
+    },
+    {
+      "epoch": 0.549800796812749,
+      "grad_norm": 0.1963123083114624,
+      "learning_rate": 4.910184225765341e-06,
+      "loss": 0.928256094455719,
+      "step": 138
+    },
+    {
+      "epoch": 0.5577689243027888,
+      "grad_norm": 0.41745948791503906,
+      "learning_rate": 4.905987845516996e-06,
+      "loss": 1.0729269981384277,
+      "step": 140
+    },
+    {
+      "epoch": 0.5657370517928287,
+      "grad_norm": 0.13186857104301453,
+      "learning_rate": 4.90169774820551e-06,
+      "loss": 0.9900568723678589,
+      "step": 142
+    },
+    {
+      "epoch": 0.5737051792828686,
+      "grad_norm": 0.10893312096595764,
+      "learning_rate": 4.897314120313996e-06,
+      "loss": 1.2397422790527344,
+      "step": 144
+    },
+    {
+      "epoch": 0.5816733067729084,
+      "grad_norm": 0.23018108308315277,
+      "learning_rate": 4.892837152391174e-06,
+      "loss": 1.406497597694397,
+      "step": 146
+    },
+    {
+      "epoch": 0.5896414342629482,
+      "grad_norm": 0.3373558819293976,
+      "learning_rate": 4.888267039043101e-06,
+      "loss": 0.912494421005249,
+      "step": 148
+    },
+    {
+      "epoch": 0.5976095617529881,
+      "grad_norm": 0.3946291506290436,
+      "learning_rate": 4.883603978924698e-06,
+      "loss": 0.9284758567810059,
+      "step": 150
+    },
+    {
+      "epoch": 0.6055776892430279,
+      "grad_norm": 0.2348312884569168,
+      "learning_rate": 4.878848174731126e-06,
+      "loss": 0.9223384261131287,
+      "step": 152
+    },
+    {
+      "epoch": 0.6135458167330677,
+      "grad_norm": 0.2850741744041443,
+      "learning_rate": 4.873999833188967e-06,
+      "loss": 1.103230595588684,
+      "step": 154
+    },
+    {
+      "epoch": 0.6215139442231076,
+      "grad_norm": 1.1040666103363037,
+      "learning_rate": 4.8690591650472455e-06,
+      "loss": 0.857952356338501,
+      "step": 156
+    },
+    {
+      "epoch": 0.6294820717131474,
+      "grad_norm": 0.19223760068416595,
+      "learning_rate": 4.864026385068259e-06,
+      "loss": 1.134782314300537,
+      "step": 158
+    },
+    {
+      "epoch": 0.6374501992031872,
+      "grad_norm": 0.247935950756073,
+      "learning_rate": 4.858901712018251e-06,
+      "loss": 0.9988402724266052,
+      "step": 160
+    },
+    {
+      "epoch": 0.6454183266932271,
+      "grad_norm": 0.08139041066169739,
+      "learning_rate": 4.8536853686578965e-06,
+      "loss": 0.7258075475692749,
+      "step": 162
+    },
+    {
+      "epoch": 0.6533864541832669,
+      "grad_norm": 0.5618287324905396,
+      "learning_rate": 4.84837758173262e-06,
+      "loss": 1.178370475769043,
+      "step": 164
+    },
+    {
+      "epoch": 0.6613545816733067,
+      "grad_norm": 0.2691766619682312,
+      "learning_rate": 4.8429785819627395e-06,
+      "loss": 0.772031843662262,
+      "step": 166
+    },
+    {
+      "epoch": 0.6693227091633466,
+      "grad_norm": 0.14576303958892822,
+      "learning_rate": 4.83748860403344e-06,
+      "loss": 1.2494381666183472,
+      "step": 168
+    },
+    {
+      "epoch": 0.6772908366533864,
+      "grad_norm": 0.5466064214706421,
+      "learning_rate": 4.831907886584566e-06,
+      "loss": 0.8462594747543335,
+      "step": 170
+    },
+    {
+      "epoch": 0.6852589641434262,
+      "grad_norm": 0.20535235106945038,
+      "learning_rate": 4.8262366722002535e-06,
+      "loss": 0.8275927901268005,
+      "step": 172
+    },
+    {
+      "epoch": 0.6932270916334662,
+      "grad_norm": 0.17850494384765625,
+      "learning_rate": 4.820475207398386e-06,
+      "loss": 1.2371994256973267,
+      "step": 174
+    },
+    {
+      "epoch": 0.701195219123506,
+      "grad_norm": 0.29729315638542175,
+      "learning_rate": 4.8146237426198735e-06,
+      "loss": 0.7060192823410034,
+      "step": 176
+    },
+    {
+      "epoch": 0.7091633466135459,
+      "grad_norm": 0.31614843010902405,
+      "learning_rate": 4.808682532217772e-06,
+      "loss": 0.8851511478424072,
+      "step": 178
+    },
+    {
+      "epoch": 0.7171314741035857,
+      "grad_norm": 0.3028443455696106,
+      "learning_rate": 4.802651834446221e-06,
+      "loss": 1.0693415403366089,
+      "step": 180
+    },
+    {
+      "epoch": 0.7250996015936255,
+      "grad_norm": 0.051578883081674576,
+      "learning_rate": 4.796531911449226e-06,
+      "loss": 0.4279974102973938,
+      "step": 182
+    },
+    {
+      "epoch": 0.7330677290836654,
+      "grad_norm": 0.11557755619287491,
+      "learning_rate": 4.790323029249252e-06,
+      "loss": 1.2791770696640015,
+      "step": 184
+    },
+    {
+      "epoch": 0.7410358565737052,
+      "grad_norm": 0.18726131319999695,
+      "learning_rate": 4.784025457735675e-06,
+      "loss": 1.2689958810806274,
+      "step": 186
+    },
+    {
+      "epoch": 0.749003984063745,
+      "grad_norm": 0.13334085047245026,
+      "learning_rate": 4.777639470653035e-06,
+      "loss": 1.0823967456817627,
+      "step": 188
+    },
+    {
+      "epoch": 0.7569721115537849,
+      "grad_norm": 0.3302866518497467,
+      "learning_rate": 4.771165345589148e-06,
+      "loss": 0.8502761125564575,
+      "step": 190
+    },
+    {
+      "epoch": 0.7649402390438247,
+      "grad_norm": 0.30881306529045105,
+      "learning_rate": 4.7646033639630355e-06,
+      "loss": 1.2255083322525024,
+      "step": 192
+    },
+    {
+      "epoch": 0.7729083665338645,
+      "grad_norm": 0.17467282712459564,
+      "learning_rate": 4.757953811012688e-06,
+      "loss": 0.8746458292007446,
+      "step": 194
+    },
+    {
+      "epoch": 0.7808764940239044,
+      "grad_norm": 0.2664046883583069,
+      "learning_rate": 4.7512169757826735e-06,
+      "loss": 0.545396625995636,
+      "step": 196
+    },
+    {
+      "epoch": 0.7888446215139442,
+      "grad_norm": 0.29917165637016296,
+      "learning_rate": 4.74439315111157e-06,
+      "loss": 0.7932103276252747,
+      "step": 198
+    },
+    {
+      "epoch": 0.796812749003984,
+      "grad_norm": 0.18899662792682648,
+      "learning_rate": 4.737482633619234e-06,
+      "loss": 0.9973336458206177,
+      "step": 200
+    },
+    {
+      "epoch": 0.8047808764940239,
+      "grad_norm": 0.08452646434307098,
+      "learning_rate": 4.730485723693912e-06,
+      "loss": 0.9312358498573303,
+      "step": 202
+    },
+    {
+      "epoch": 0.8127490039840638,
+      "grad_norm": 0.2051236629486084,
+      "learning_rate": 4.723402725479177e-06,
+      "loss": 1.5546748638153076,
+      "step": 204
+    },
+    {
+      "epoch": 0.8207171314741036,
+      "grad_norm": 0.2771088778972626,
+      "learning_rate": 4.71623394686071e-06,
+      "loss": 1.2395048141479492,
+      "step": 206
+    },
+    {
+      "epoch": 0.8286852589641435,
+      "grad_norm": 0.09508787095546722,
+      "learning_rate": 4.708979699452923e-06,
+      "loss": 0.5828080773353577,
+      "step": 208
+    },
+    {
+      "epoch": 0.8366533864541833,
+      "grad_norm": 0.2051192969083786,
+      "learning_rate": 4.701640298585405e-06,
+      "loss": 1.2725114822387695,
+      "step": 210
+    },
+    {
+      "epoch": 0.8446215139442231,
+      "grad_norm": 0.12022526562213898,
+      "learning_rate": 4.694216063289219e-06,
+      "loss": 1.2300056219100952,
+      "step": 212
+    },
+    {
+      "epoch": 0.852589641434263,
+      "grad_norm": 0.13225895166397095,
+      "learning_rate": 4.686707316283033e-06,
+      "loss": 0.9437692165374756,
+      "step": 214
+    },
+    {
+      "epoch": 0.8605577689243028,
+      "grad_norm": 0.23339490592479706,
+      "learning_rate": 4.679114383959095e-06,
+      "loss": 1.4291656017303467,
+      "step": 216
+    },
+    {
+      "epoch": 0.8685258964143426,
+      "grad_norm": 0.23099103569984436,
+      "learning_rate": 4.6714375963690415e-06,
+      "loss": 0.8413586616516113,
+      "step": 218
+    },
+    {
+      "epoch": 0.8764940239043825,
+      "grad_norm": 0.11677176505327225,
+      "learning_rate": 4.6636772872095545e-06,
+      "loss": 1.2480651140213013,
+      "step": 220
+    },
+    {
+      "epoch": 0.8844621513944223,
+      "grad_norm": 0.4693072736263275,
+      "learning_rate": 4.6558337938078505e-06,
+      "loss": 1.491629719734192,
+      "step": 222
+    },
+    {
+      "epoch": 0.8924302788844621,
+      "grad_norm": 0.5661599636077881,
+      "learning_rate": 4.647907457107025e-06,
+      "loss": 0.6979908347129822,
+      "step": 224
+    },
+    {
+      "epoch": 0.900398406374502,
+      "grad_norm": 0.21899700164794922,
+      "learning_rate": 4.639898621651226e-06,
+      "loss": 0.9708130955696106,
+      "step": 226
+    },
+    {
+      "epoch": 0.9083665338645418,
+      "grad_norm": 0.3328002095222473,
+      "learning_rate": 4.631807635570681e-06,
+      "loss": 1.0735762119293213,
+      "step": 228
+    },
+    {
+      "epoch": 0.9163346613545816,
+      "grad_norm": 0.18838559091091156,
+      "learning_rate": 4.623634850566563e-06,
+      "loss": 1.17432701587677,
+      "step": 230
+    },
+    {
+      "epoch": 0.9243027888446215,
+      "grad_norm": 0.34923747181892395,
+      "learning_rate": 4.615380621895701e-06,
+      "loss": 1.3521753549575806,
+      "step": 232
+    },
+    {
+      "epoch": 0.9322709163346613,
+      "grad_norm": 0.10172160714864731,
+      "learning_rate": 4.607045308355141e-06,
+      "loss": 0.7744335532188416,
+      "step": 234
+    },
+    {
+      "epoch": 0.9402390438247012,
+      "grad_norm": 0.2071087658405304,
+      "learning_rate": 4.598629272266549e-06,
+      "loss": 0.7924075126647949,
+      "step": 236
+    },
+    {
+      "epoch": 0.9482071713147411,
+      "grad_norm": 0.134637251496315,
+      "learning_rate": 4.590132879460458e-06,
+      "loss": 0.406107097864151,
+      "step": 238
+    },
+    {
+      "epoch": 0.9561752988047809,
+      "grad_norm": 0.161122128367424,
+      "learning_rate": 4.5815564992603705e-06,
+      "loss": 0.8294285535812378,
+      "step": 240
+    },
+    {
+      "epoch": 0.9641434262948207,
+      "grad_norm": 0.19272178411483765,
+      "learning_rate": 4.572900504466702e-06,
+      "loss": 1.1219251155853271,
+      "step": 242
+    },
+    {
+      "epoch": 0.9721115537848606,
+      "grad_norm": 0.1935759037733078,
+      "learning_rate": 4.564165271340577e-06,
+      "loss": 1.2249534130096436,
+      "step": 244
+    },
+    {
+      "epoch": 0.9800796812749004,
+      "grad_norm": 0.1313743144273758,
+      "learning_rate": 4.555351179587472e-06,
+      "loss": 0.9394630789756775,
+      "step": 246
+    },
+    {
+      "epoch": 0.9880478087649402,
+      "grad_norm": 0.09965435415506363,
+      "learning_rate": 4.546458612340714e-06,
+      "loss": 0.9602974653244019,
+      "step": 248
+    },
+    {
+      "epoch": 0.9960159362549801,
+      "grad_norm": 0.1887800693511963,
+      "learning_rate": 4.537487956144821e-06,
+      "loss": 0.9729372262954712,
+      "step": 250
+    },
+    {
+      "epoch": 1.00398406374502,
+      "grad_norm": 0.6515352725982666,
+      "learning_rate": 4.528439600938706e-06,
+      "loss": 0.8933233618736267,
+      "step": 252
+    },
+    {
+      "epoch": 1.0119521912350598,
+      "grad_norm": 0.2628687620162964,
+      "learning_rate": 4.519313940038721e-06,
+      "loss": 0.6970143914222717,
+      "step": 254
+    },
+    {
+      "epoch": 1.0199203187250996,
+      "grad_norm": 0.16754353046417236,
+      "learning_rate": 4.510111370121563e-06,
+      "loss": 1.192719578742981,
+      "step": 256
+    },
+    {
+      "epoch": 1.0278884462151394,
+      "grad_norm": 0.17667120695114136,
+      "learning_rate": 4.500832291207033e-06,
+      "loss": 1.1800732612609863,
+      "step": 258
+    },
+    {
+      "epoch": 1.0358565737051793,
+      "grad_norm": 0.17618392407894135,
+      "learning_rate": 4.491477106640646e-06,
+      "loss": 1.107790470123291,
+      "step": 260
+    },
+    {
+      "epoch": 1.043824701195219,
+      "grad_norm": 0.19453424215316772,
+      "learning_rate": 4.482046223076092e-06,
+      "loss": 1.0330584049224854,
+      "step": 262
+    },
+    {
+      "epoch": 1.051792828685259,
+      "grad_norm": 0.16134338080883026,
+      "learning_rate": 4.4725400504575725e-06,
+      "loss": 1.0822701454162598,
+      "step": 264
+    },
+    {
+      "epoch": 1.0597609561752988,
+      "grad_norm": 0.3839372396469116,
+      "learning_rate": 4.462959002001971e-06,
+      "loss": 0.6190903186798096,
+      "step": 266
+    },
+    {
+      "epoch": 1.0677290836653386,
+      "grad_norm": 0.10365067422389984,
+      "learning_rate": 4.453303494180895e-06,
+      "loss": 0.9371901154518127,
+      "step": 268
+    },
+    {
+      "epoch": 1.0756972111553784,
+      "grad_norm": 0.6638212203979492,
+      "learning_rate": 4.443573946702573e-06,
+      "loss": 0.5588272213935852,
+      "step": 270
+    },
+    {
+      "epoch": 1.0836653386454183,
+      "grad_norm": 0.1912950575351715,
+      "learning_rate": 4.433770782493606e-06,
+      "loss": 1.2535881996154785,
+      "step": 272
+    },
+    {
+      "epoch": 1.091633466135458,
+      "grad_norm": 0.2721278965473175,
+      "learning_rate": 4.42389442768059e-06,
+      "loss": 0.5842412114143372,
+      "step": 274
+    },
+    {
+      "epoch": 1.099601593625498,
+      "grad_norm": 0.15831194818019867,
+      "learning_rate": 4.413945311571586e-06,
+      "loss": 0.9588902592658997,
+      "step": 276
+    },
+    {
+      "epoch": 1.1075697211155378,
+      "grad_norm": 0.159891277551651,
+      "learning_rate": 4.40392386663747e-06,
+      "loss": 0.837569534778595,
+      "step": 278
+    },
+    {
+      "epoch": 1.1155378486055776,
+      "grad_norm": 0.2322358787059784,
+      "learning_rate": 4.393830528493118e-06,
+      "loss": 1.0724090337753296,
+      "step": 280
+    },
+    {
+      "epoch": 1.1235059760956174,
+      "grad_norm": 0.16406196355819702,
+      "learning_rate": 4.3836657358784875e-06,
+      "loss": 1.2008377313613892,
+      "step": 282
+    },
+    {
+      "epoch": 1.1314741035856573,
+      "grad_norm": 0.2790522873401642,
+      "learning_rate": 4.373429930639534e-06,
+      "loss": 1.2208435535430908,
+      "step": 284
+    },
+    {
+      "epoch": 1.139442231075697,
+      "grad_norm": 0.19494254887104034,
+      "learning_rate": 4.3631235577090085e-06,
+      "loss": 0.27236077189445496,
+      "step": 286
+    },
+    {
+      "epoch": 1.1474103585657371,
+      "grad_norm": 0.10534805804491043,
+      "learning_rate": 4.35274706508712e-06,
+      "loss": 0.7416235208511353,
+      "step": 288
+    },
+    {
+      "epoch": 1.155378486055777,
+      "grad_norm": 0.7741219401359558,
+      "learning_rate": 4.342300903822057e-06,
+      "loss": 0.9409246444702148,
+      "step": 290
+    },
+    {
+      "epoch": 1.1633466135458168,
+      "grad_norm": 0.17461933195590973,
+      "learning_rate": 4.331785527990384e-06,
+      "loss": 0.5023335814476013,
+      "step": 292
+    },
+    {
+      "epoch": 1.1713147410358566,
+      "grad_norm": 0.14229869842529297,
+      "learning_rate": 4.3212013946773045e-06,
+      "loss": 1.2262612581253052,
+      "step": 294
+    },
+    {
+      "epoch": 1.1792828685258965,
+      "grad_norm": 0.145178884267807,
+      "learning_rate": 4.31054896395679e-06,
+      "loss": 1.2516542673110962,
+      "step": 296
+    },
+    {
+      "epoch": 1.1872509960159363,
+      "grad_norm": 0.22644661366939545,
+      "learning_rate": 4.299828698871581e-06,
+      "loss": 1.4002199172973633,
+      "step": 298
+    },
+    {
+      "epoch": 1.1952191235059761,
+      "grad_norm": 0.20929889380931854,
+      "learning_rate": 4.289041065413062e-06,
+      "loss": 0.5912055969238281,
+      "step": 300
+    },
+    {
+      "epoch": 1.203187250996016,
+      "grad_norm": 0.1715276539325714,
+      "learning_rate": 4.278186532501003e-06,
+      "loss": 0.9038018584251404,
+      "step": 302
+    },
+    {
+      "epoch": 1.2111553784860558,
+      "grad_norm": 0.2748669981956482,
+      "learning_rate": 4.267265571963179e-06,
+      "loss": 0.5407243371009827,
+      "step": 304
+    },
+    {
+      "epoch": 1.2191235059760956,
+      "grad_norm": 0.14455609023571014,
+      "learning_rate": 4.256278658514858e-06,
+      "loss": 1.202244520187378,
+      "step": 306
+    },
+    {
+      "epoch": 1.2270916334661355,
+      "grad_norm": 0.7603815197944641,
+      "learning_rate": 4.245226269738165e-06,
+      "loss": 1.0239896774291992,
+      "step": 308
+    },
+    {
+      "epoch": 1.2350597609561753,
+      "grad_norm": 0.3512886166572571,
+      "learning_rate": 4.23410888606133e-06,
+      "loss": 1.1498864889144897,
+      "step": 310
+    },
+    {
+      "epoch": 1.2430278884462151,
+      "grad_norm": 0.32941585779190063,
+      "learning_rate": 4.222926990737793e-06,
+      "loss": 1.0864653587341309,
+      "step": 312
+    },
+    {
+      "epoch": 1.250996015936255,
+      "grad_norm": 0.4355418086051941,
+      "learning_rate": 4.211681069825207e-06,
+      "loss": 0.8547599911689758,
+      "step": 314
+    },
+    {
+      "epoch": 1.2589641434262948,
+      "grad_norm": 0.1957080066204071,
+      "learning_rate": 4.200371612164305e-06,
+      "loss": 0.7279231548309326,
+      "step": 316
+    },
+    {
+      "epoch": 1.2669322709163346,
+      "grad_norm": 0.31115415692329407,
+      "learning_rate": 4.1889991093576555e-06,
+      "loss": 0.6071080565452576,
+      "step": 318
+    },
+    {
+      "epoch": 1.2749003984063745,
+      "grad_norm": 0.27555859088897705,
+      "learning_rate": 4.177564055748284e-06,
+      "loss": 0.5027056336402893,
+      "step": 320
+    },
+    {
+      "epoch": 1.2828685258964143,
+      "grad_norm": 0.366533100605011,
+      "learning_rate": 4.166066948398199e-06,
+      "loss": 1.0088260173797607,
+      "step": 322
+    },
+    {
+      "epoch": 1.2908366533864541,
+      "grad_norm": 0.19128911197185516,
+      "learning_rate": 4.154508287066772e-06,
+      "loss": 0.9387359619140625,
+      "step": 324
+    },
+    {
+      "epoch": 1.298804780876494,
+      "grad_norm": 1.9903514385223389,
+      "learning_rate": 4.142888574189025e-06,
+      "loss": 1.3278898000717163,
+      "step": 326
+    },
+    {
+      "epoch": 1.3067729083665338,
+      "grad_norm": 0.16312150657176971,
+      "learning_rate": 4.131208314853778e-06,
+      "loss": 0.7040650248527527,
+      "step": 328
+    },
+    {
+      "epoch": 1.3147410358565736,
+      "grad_norm": 0.20557750761508942,
+      "learning_rate": 4.119468016781707e-06,
+      "loss": 0.9129243493080139,
+      "step": 330
+    },
+    {
+      "epoch": 1.3227091633466135,
+      "grad_norm": 0.37182188034057617,
+      "learning_rate": 4.107668190303266e-06,
+      "loss": 0.9323933720588684,
+      "step": 332
+    },
+    {
+      "epoch": 1.3306772908366533,
+      "grad_norm": 0.3476126492023468,
+      "learning_rate": 4.095809348336506e-06,
+      "loss": 0.6508227586746216,
+      "step": 334
+    },
+    {
+      "epoch": 1.3386454183266931,
+      "grad_norm": 0.44167131185531616,
+      "learning_rate": 4.083892006364777e-06,
+      "loss": 0.9058379530906677,
+      "step": 336
+    },
+    {
+      "epoch": 1.3466135458167332,
+      "grad_norm": 0.25710543990135193,
+      "learning_rate": 4.071916682414327e-06,
+      "loss": 1.1918373107910156,
+      "step": 338
+    },
+    {
+      "epoch": 1.354581673306773,
+      "grad_norm": 0.19701232016086578,
+      "learning_rate": 4.059883897031777e-06,
+      "loss": 1.1804696321487427,
+      "step": 340
+    },
+    {
+      "epoch": 1.3625498007968129,
+      "grad_norm": 0.1606234312057495,
+      "learning_rate": 4.0477941732615e-06,
+      "loss": 1.2122278213500977,
+      "step": 342
+    },
+    {
+      "epoch": 1.3705179282868527,
+      "grad_norm": 0.29800689220428467,
+      "learning_rate": 4.035648036622879e-06,
+      "loss": 1.1901203393936157,
+      "step": 344
+    },
+    {
+      "epoch": 1.3784860557768925,
+      "grad_norm": 0.12151093780994415,
+      "learning_rate": 4.023446015087469e-06,
+      "loss": 0.7064657211303711,
+      "step": 346
+    },
+    {
+      "epoch": 1.3864541832669324,
+      "grad_norm": 0.15453097224235535,
+      "learning_rate": 4.0111886390560446e-06,
+      "loss": 1.1941863298416138,
+      "step": 348
+    },
+    {
+      "epoch": 1.3944223107569722,
+      "grad_norm": 0.22483372688293457,
+      "learning_rate": 3.998876441335544e-06,
+      "loss": 0.7265820503234863,
+      "step": 350
+    },
+    {
+      "epoch": 1.402390438247012,
+      "grad_norm": 0.14736157655715942,
+      "learning_rate": 3.98650995711591e-06,
+      "loss": 1.2339738607406616,
+      "step": 352
+    },
+    {
+      "epoch": 1.4103585657370519,
+      "grad_norm": 0.20219145715236664,
+      "learning_rate": 3.974089723946826e-06,
+      "loss": 1.1635406017303467,
+      "step": 354
+    },
+    {
+      "epoch": 1.4183266932270917,
+      "grad_norm": 0.233228400349617,
+      "learning_rate": 3.9616162817143465e-06,
+      "loss": 0.9164896607398987,
+      "step": 356
+    },
+    {
+      "epoch": 1.4262948207171315,
+      "grad_norm": 0.23091474175453186,
+      "learning_rate": 3.949090172617437e-06,
+      "loss": 1.1780691146850586,
+      "step": 358
+    },
+    {
+      "epoch": 1.4342629482071714,
+      "grad_norm": 0.16404017806053162,
+      "learning_rate": 3.936511941144397e-06,
+      "loss": 0.6284229159355164,
+      "step": 360
+    },
+    {
+      "epoch": 1.4422310756972112,
+      "grad_norm": 0.131036639213562,
+      "learning_rate": 3.9238821340491936e-06,
+      "loss": 0.760465145111084,
+      "step": 362
+    },
+    {
+      "epoch": 1.450199203187251,
+      "grad_norm": 0.2481812983751297,
+      "learning_rate": 3.911201300327703e-06,
+      "loss": 1.0749475955963135,
+      "step": 364
+    },
+    {
+      "epoch": 1.4581673306772909,
+      "grad_norm": 0.15491396188735962,
+      "learning_rate": 3.898469991193835e-06,
+      "loss": 1.1622352600097656,
+      "step": 366
+    },
+    {
+      "epoch": 1.4661354581673307,
+      "grad_norm": 0.24048905074596405,
+      "learning_rate": 3.885688760055581e-06,
+      "loss": 0.6309763193130493,
+      "step": 368
+    },
+    {
+      "epoch": 1.4741035856573705,
+      "grad_norm": 0.25780460238456726,
+      "learning_rate": 3.872858162490956e-06,
+      "loss": 0.721203088760376,
+      "step": 370
+    },
+    {
+      "epoch": 1.4820717131474104,
+      "grad_norm": 0.3392886221408844,
+      "learning_rate": 3.859978756223845e-06,
+      "loss": 1.2578173875808716,
+      "step": 372
+    },
+    {
+      "epoch": 1.4900398406374502,
+      "grad_norm": 0.18764694035053253,
+      "learning_rate": 3.847051101099768e-06,
+      "loss": 1.1831198930740356,
+      "step": 374
+    },
+    {
+      "epoch": 1.49800796812749,
+      "grad_norm": 0.13412053883075714,
+      "learning_rate": 3.834075759061535e-06,
+      "loss": 1.175061583518982,
+      "step": 376
+    },
+    {
+      "epoch": 1.5059760956175299,
+      "grad_norm": 0.800938606262207,
+      "learning_rate": 3.8210532941248255e-06,
+      "loss": 0.950330376625061,
+      "step": 378
+    },
+    {
+      "epoch": 1.5139442231075697,
+      "grad_norm": 0.1650896966457367,
+      "learning_rate": 3.8079842723536688e-06,
+      "loss": 1.1499178409576416,
+      "step": 380
+    },
+    {
+      "epoch": 1.5219123505976095,
+      "grad_norm": 0.11898193508386612,
+      "learning_rate": 3.7948692618358424e-06,
+      "loss": 0.7940158247947693,
+      "step": 382
+    },
+    {
+      "epoch": 1.5298804780876494,
+      "grad_norm": 0.23134896159172058,
+      "learning_rate": 3.781708832658169e-06,
+      "loss": 0.782852053642273,
+      "step": 384
+    },
+    {
+      "epoch": 1.5378486055776892,
+      "grad_norm": 0.24510852992534637,
+      "learning_rate": 3.7685035568817492e-06,
+      "loss": 1.0700899362564087,
+      "step": 386
+    },
+    {
+      "epoch": 1.545816733067729,
+      "grad_norm": 0.3537372350692749,
+      "learning_rate": 3.7552540085170826e-06,
+      "loss": 1.190940022468567,
+      "step": 388
+    },
+    {
+      "epoch": 1.5537848605577689,
+      "grad_norm": 0.22087712585926056,
+      "learning_rate": 3.7419607634991235e-06,
+      "loss": 0.5493913292884827,
+      "step": 390
+    },
+    {
+      "epoch": 1.5617529880478087,
+      "grad_norm": 0.3640039563179016,
+      "learning_rate": 3.7286243996622432e-06,
+      "loss": 0.589182436466217,
+      "step": 392
+    },
+    {
+      "epoch": 1.5697211155378485,
+      "grad_norm": 0.42739778757095337,
+      "learning_rate": 3.715245496715116e-06,
+      "loss": 1.1656575202941895,
+      "step": 394
+    },
+    {
+      "epoch": 1.5776892430278884,
+      "grad_norm": 0.6680868268013,
+      "learning_rate": 3.7018246362155134e-06,
+      "loss": 1.0725603103637695,
+      "step": 396
+    },
+    {
+      "epoch": 1.5856573705179282,
+      "grad_norm": 0.17477107048034668,
+      "learning_rate": 3.6883624015450344e-06,
+      "loss": 1.1672534942626953,
+      "step": 398
+    },
+    {
+      "epoch": 1.593625498007968,
+      "grad_norm": 0.2545846998691559,
+      "learning_rate": 3.6748593778837367e-06,
+      "loss": 0.5984270572662354,
+      "step": 400
+    },
+    {
+      "epoch": 1.6015936254980079,
+      "grad_norm": 0.1376497447490692,
+      "learning_rate": 3.661316152184708e-06,
+      "loss": 1.1501036882400513,
+      "step": 402
+    },
+    {
+      "epoch": 1.6095617529880477,
+      "grad_norm": 0.3141114413738251,
+      "learning_rate": 3.6477333131485483e-06,
+      "loss": 1.3023042678833008,
+      "step": 404
+    },
+    {
+      "epoch": 1.6175298804780875,
+      "grad_norm": 0.15489861369132996,
+      "learning_rate": 3.634111451197781e-06,
+      "loss": 0.7329400777816772,
+      "step": 406
+    },
+    {
+      "epoch": 1.6254980079681274,
+      "grad_norm": 0.15180833637714386,
+      "learning_rate": 3.6204511584511865e-06,
+      "loss": 0.5869722962379456,
+      "step": 408
+    },
+    {
+      "epoch": 1.6334661354581672,
+      "grad_norm": 0.5213863849639893,
+      "learning_rate": 3.606753028698067e-06,
+      "loss": 0.8130780458450317,
+      "step": 410
+    },
+    {
+      "epoch": 1.641434262948207,
+      "grad_norm": 0.6388369202613831,
+      "learning_rate": 3.593017657372436e-06,
+      "loss": 0.5190494656562805,
+      "step": 412
+    },
+    {
+      "epoch": 1.6494023904382469,
+      "grad_norm": 0.12885528802871704,
+      "learning_rate": 3.5792456415271288e-06,
+      "loss": 0.8747182488441467,
+      "step": 414
+    },
+    {
+      "epoch": 1.6573705179282867,
+      "grad_norm": 0.4794620871543884,
+      "learning_rate": 3.565437579807858e-06,
+      "loss": 0.7429513335227966,
+      "step": 416
+    },
+    {
+      "epoch": 1.6653386454183265,
+      "grad_norm": 0.2635369896888733,
+      "learning_rate": 3.551594072427188e-06,
+      "loss": 1.0676034688949585,
+      "step": 418
+    },
+    {
+      "epoch": 1.6733067729083664,
+      "grad_norm": 0.2294226586818695,
+      "learning_rate": 3.537715721138444e-06,
+      "loss": 1.163661003112793,
+      "step": 420
+    },
+    {
+      "epoch": 1.6812749003984062,
+      "grad_norm": 0.26409414410591125,
+      "learning_rate": 3.523803129209554e-06,
+      "loss": 1.0647904872894287,
+      "step": 422
+    },
+    {
+      "epoch": 1.6892430278884463,
+      "grad_norm": 0.24496956169605255,
+      "learning_rate": 3.5098569013968314e-06,
+      "loss": 0.6253319382667542,
+      "step": 424
+    },
+    {
+      "epoch": 1.697211155378486,
+      "grad_norm": 0.14838750660419464,
+      "learning_rate": 3.49587764391868e-06,
+      "loss": 1.2408252954483032,
+      "step": 426
+    },
+    {
+      "epoch": 1.705179282868526,
+      "grad_norm": 0.16086408495903015,
+      "learning_rate": 3.4818659644292473e-06,
+      "loss": 1.1767014265060425,
+      "step": 428
+    },
+    {
+      "epoch": 1.7131474103585658,
+      "grad_norm": 0.1459408402442932,
+      "learning_rate": 3.4678224719920085e-06,
+      "loss": 0.7547453045845032,
+      "step": 430
+    },
+    {
+      "epoch": 1.7211155378486056,
+      "grad_norm": 0.2707553505897522,
+      "learning_rate": 3.4537477770532933e-06,
+      "loss": 0.9470460414886475,
+      "step": 432
+    },
+    {
+      "epoch": 1.7290836653386454,
+      "grad_norm": 0.38563284277915955,
+      "learning_rate": 3.4396424914157527e-06,
+      "loss": 0.6457641124725342,
+      "step": 434
+    },
+    {
+      "epoch": 1.7370517928286853,
+      "grad_norm": 0.29641658067703247,
+      "learning_rate": 3.4255072282117574e-06,
+      "loss": 0.3550601601600647,
+      "step": 436
+    },
+    {
+      "epoch": 1.745019920318725,
+      "grad_norm": 0.22587662935256958,
+      "learning_rate": 3.411342601876754e-06,
+      "loss": 0.7444441914558411,
+      "step": 438
+    },
+    {
+      "epoch": 1.752988047808765,
+      "grad_norm": 0.19855016469955444,
+      "learning_rate": 3.3971492281225554e-06,
+      "loss": 0.6208574175834656,
+      "step": 440
+    },
+    {
+      "epoch": 1.7609561752988048,
+      "grad_norm": 0.29874321818351746,
+      "learning_rate": 3.382927723910572e-06,
+      "loss": 1.5587701797485352,
+      "step": 442
+    },
+    {
+      "epoch": 1.7689243027888446,
+      "grad_norm": 0.20580004155635834,
+      "learning_rate": 3.3686787074249966e-06,
+      "loss": 0.8196008801460266,
+      "step": 444
+    },
+    {
+      "epoch": 1.7768924302788844,
+      "grad_norm": 0.11870943754911423,
+      "learning_rate": 3.354402798045935e-06,
+      "loss": 1.0533963441848755,
+      "step": 446
+    },
+    {
+      "epoch": 1.7848605577689243,
+      "grad_norm": 0.15005037188529968,
+      "learning_rate": 3.340100616322479e-06,
+      "loss": 1.1275280714035034,
+      "step": 448
+    },
+    {
+      "epoch": 1.792828685258964,
+      "grad_norm": 0.09767521917819977,
+      "learning_rate": 3.3257727839457356e-06,
+      "loss": 0.8568601012229919,
+      "step": 450
+    },
+    {
+      "epoch": 1.800796812749004,
+      "grad_norm": 0.13437511026859283,
+      "learning_rate": 3.311419923721798e-06,
+      "loss": 0.5592570900917053,
+      "step": 452
+    },
+    {
+      "epoch": 1.8087649402390438,
+      "grad_norm": 0.528956413269043,
+      "learning_rate": 3.29704265954468e-06,
+      "loss": 1.0183025598526,
+      "step": 454
+    },
+    {
+      "epoch": 1.8167330677290838,
+      "grad_norm": 0.359781414270401,
+      "learning_rate": 3.282641616369193e-06,
+      "loss": 0.6656947731971741,
+      "step": 456
+    },
+    {
+      "epoch": 1.8247011952191237,
+      "grad_norm": 0.2265646606683731,
+      "learning_rate": 3.2682174201837773e-06,
+      "loss": 1.0649449825286865,
+      "step": 458
+    },
+    {
+      "epoch": 1.8326693227091635,
+      "grad_norm": 0.24529722332954407,
+      "learning_rate": 3.2537706979832955e-06,
+      "loss": 0.735980749130249,
+      "step": 460
+    },
+    {
+      "epoch": 1.8406374501992033,
+      "grad_norm": 0.10968779027462006,
+      "learning_rate": 3.23930207774178e-06,
+      "loss": 0.7897536754608154,
+      "step": 462
+    },
+    {
+      "epoch": 1.8486055776892432,
+      "grad_norm": 0.07297369092702866,
+      "learning_rate": 3.2248121883851296e-06,
+      "loss": 0.12234770506620407,
+      "step": 464
+    },
+    {
+      "epoch": 1.856573705179283,
+      "grad_norm": 0.132096529006958,
+      "learning_rate": 3.210301659763777e-06,
+      "loss": 1.1873613595962524,
+      "step": 466
+    },
+    {
+      "epoch": 1.8645418326693228,
+      "grad_norm": 0.24179653823375702,
+      "learning_rate": 3.1957711226253073e-06,
+      "loss": 0.5585373640060425,
+      "step": 468
+    },
+    {
+      "epoch": 1.8725099601593627,
+      "grad_norm": 0.2070511281490326,
+      "learning_rate": 3.181221208587041e-06,
+      "loss": 1.1847221851348877,
+      "step": 470
+    },
+    {
+      "epoch": 1.8804780876494025,
+      "grad_norm": 0.2511274814605713,
+      "learning_rate": 3.16665255010858e-06,
+      "loss": 1.0381547212600708,
+      "step": 472
+    },
+    {
+      "epoch": 1.8884462151394423,
+      "grad_norm": 0.5230444073677063,
+      "learning_rate": 3.152065780464311e-06,
+      "loss": 1.2714117765426636,
+      "step": 474
+    },
+    {
+      "epoch": 1.8964143426294822,
+      "grad_norm": 0.09674343466758728,
+      "learning_rate": 3.1374615337158874e-06,
+      "loss": 0.9738749265670776,
+      "step": 476
+    },
+    {
+      "epoch": 1.904382470119522,
+      "grad_norm": 0.16097377240657806,
+      "learning_rate": 3.1228404446846577e-06,
+      "loss": 0.8900617361068726,
+      "step": 478
+    },
+    {
+      "epoch": 1.9123505976095618,
+      "grad_norm": 0.12337128072977066,
+      "learning_rate": 3.1082031489240775e-06,
+      "loss": 0.5881850719451904,
+      "step": 480
+    },
+    {
+      "epoch": 1.9203187250996017,
+      "grad_norm": 0.22677840292453766,
+      "learning_rate": 3.093550282692081e-06,
+      "loss": 0.7393006682395935,
+      "step": 482
+    },
+    {
+      "epoch": 1.9282868525896415,
+      "grad_norm": 0.14126428961753845,
+      "learning_rate": 3.0788824829234206e-06,
+      "loss": 1.1454389095306396,
+      "step": 484
+    },
+    {
+      "epoch": 1.9362549800796813,
+      "grad_norm": 0.32119759917259216,
+      "learning_rate": 3.064200387201987e-06,
+      "loss": 1.1732426881790161,
+      "step": 486
+    },
+    {
+      "epoch": 1.9442231075697212,
+      "grad_norm": 0.11168768256902695,
+      "learning_rate": 3.0495046337330917e-06,
+      "loss": 1.1405354738235474,
+      "step": 488
+    },
+    {
+      "epoch": 1.952191235059761,
+      "grad_norm": 0.09070200473070145,
+      "learning_rate": 3.034795861315723e-06,
+      "loss": 1.1537675857543945,
+      "step": 490
+    },
+    {
+      "epoch": 1.9601593625498008,
+      "grad_norm": 0.14500442147254944,
+      "learning_rate": 3.020074709314781e-06,
+      "loss": 0.8191499710083008,
+      "step": 492
+    },
+    {
+      "epoch": 1.9681274900398407,
+      "grad_norm": 1.1981549263000488,
+      "learning_rate": 3.0053418176332844e-06,
+      "loss": 0.6050922870635986,
+      "step": 494
+    },
+    {
+      "epoch": 1.9760956175298805,
+      "grad_norm": 0.09703997522592545,
+      "learning_rate": 2.9905978266845565e-06,
+      "loss": 1.1263906955718994,
+      "step": 496
+    },
+    {
+      "epoch": 1.9840637450199203,
+      "grad_norm": 0.2982192039489746,
+      "learning_rate": 2.975843377364385e-06,
+      "loss": 0.8881535530090332,
+      "step": 498
+    },
+    {
+      "epoch": 1.9920318725099602,
+      "grad_norm": 0.09076780080795288,
+      "learning_rate": 2.961079111023167e-06,
+      "loss": 1.1868828535079956,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.15667445957660675,
+      "learning_rate": 2.9463056694380263e-06,
+      "loss": 0.6327648162841797,
+      "step": 502
+    },
+    {
+      "epoch": 2.00796812749004,
+      "grad_norm": 0.11475303024053574,
+      "learning_rate": 2.9315236947849213e-06,
+      "loss": 0.674118161201477,
+      "step": 504
+    },
+    {
+      "epoch": 2.0159362549800797,
+      "grad_norm": 0.2793906629085541,
+      "learning_rate": 2.916733829610726e-06,
+      "loss": 0.6947035193443298,
+      "step": 506
+    },
+    {
+      "epoch": 2.0239043824701195,
+      "grad_norm": 0.0297082532197237,
+      "learning_rate": 2.901936716805302e-06,
+      "loss": 0.5822932720184326,
+      "step": 508
+    },
+    {
+      "epoch": 2.0318725099601593,
+      "grad_norm": 0.19081753492355347,
+      "learning_rate": 2.887132999573554e-06,
+      "loss": 0.9074264764785767,
+      "step": 510
+    },
+    {
+      "epoch": 2.039840637450199,
+      "grad_norm": 0.20524288713932037,
+      "learning_rate": 2.8723233214074675e-06,
+      "loss": 0.7811470031738281,
+      "step": 512
+    },
+    {
+      "epoch": 2.047808764940239,
+      "grad_norm": 0.5949530601501465,
+      "learning_rate": 2.857508326058142e-06,
+      "loss": 0.294197142124176,
+      "step": 514
+    },
+    {
+      "epoch": 2.055776892430279,
+      "grad_norm": 0.11781960725784302,
+      "learning_rate": 2.842688657507805e-06,
+      "loss": 0.9331067204475403,
+      "step": 516
+    },
+    {
+      "epoch": 2.0637450199203187,
+      "grad_norm": 0.198091521859169,
+      "learning_rate": 2.8278649599418194e-06,
+      "loss": 0.5611523389816284,
+      "step": 518
+    },
+    {
+      "epoch": 2.0717131474103585,
+      "grad_norm": 0.17517666518688202,
+      "learning_rate": 2.813037877720683e-06,
+      "loss": 0.9161994457244873,
+      "step": 520
+    },
+    {
+      "epoch": 2.0796812749003983,
+      "grad_norm": 0.17486606538295746,
+      "learning_rate": 2.798208055352017e-06,
+      "loss": 1.0747190713882446,
+      "step": 522
+    },
+    {
+      "epoch": 2.087649402390438,
+      "grad_norm": 0.3806056082248688,
+      "learning_rate": 2.7833761374625546e-06,
+      "loss": 1.0060160160064697,
+      "step": 524
+    },
+    {
+      "epoch": 2.095617529880478,
+      "grad_norm": 0.21008682250976562,
+      "learning_rate": 2.7685427687701172e-06,
+      "loss": 0.44929295778274536,
+      "step": 526
+    },
+    {
+      "epoch": 2.103585657370518,
+      "grad_norm": 0.19024591147899628,
+      "learning_rate": 2.753708594055588e-06,
+      "loss": 0.8603243827819824,
+      "step": 528
+    },
+    {
+      "epoch": 2.1115537848605577,
+      "grad_norm": 0.09967364370822906,
+      "learning_rate": 2.7388742581348893e-06,
+      "loss": 0.7523764371871948,
+      "step": 530
+    },
+    {
+      "epoch": 2.1195219123505975,
+      "grad_norm": 0.26462322473526,
+      "learning_rate": 2.7240404058309487e-06,
+      "loss": 0.26911529898643494,
+      "step": 532
+    },
+    {
+      "epoch": 2.1274900398406373,
+      "grad_norm": 0.3541204333305359,
+      "learning_rate": 2.709207681945673e-06,
+      "loss": 0.8198036551475525,
+      "step": 534
+    },
+    {
+      "epoch": 2.135458167330677,
+      "grad_norm": 0.3390915095806122,
+      "learning_rate": 2.6943767312319167e-06,
+      "loss": 0.6654301285743713,
+      "step": 536
+    },
+    {
+      "epoch": 2.143426294820717,
+      "grad_norm": 0.2651524543762207,
+      "learning_rate": 2.679548198365462e-06,
+      "loss": 0.7026398181915283,
+      "step": 538
+    },
+    {
+      "epoch": 2.151394422310757,
+      "grad_norm": 0.23493485152721405,
+      "learning_rate": 2.6647227279169873e-06,
+      "loss": 0.9859932065010071,
+      "step": 540
+    },
+    {
+      "epoch": 2.1593625498007967,
+      "grad_norm": 0.23338450491428375,
+      "learning_rate": 2.6499009643240527e-06,
+      "loss": 0.9928818941116333,
+      "step": 542
+    },
+    {
+      "epoch": 2.1673306772908365,
+      "grad_norm": 0.15555372834205627,
+      "learning_rate": 2.6350835518630907e-06,
+      "loss": 0.8929111957550049,
+      "step": 544
+    },
+    {
+      "epoch": 2.1752988047808763,
+      "grad_norm": 0.19031359255313873,
+      "learning_rate": 2.620271134621397e-06,
+      "loss": 0.6852139234542847,
+      "step": 546
+    },
+    {
+      "epoch": 2.183266932270916,
+      "grad_norm": 0.15254168212413788,
+      "learning_rate": 2.605464356469131e-06,
+      "loss": 1.056442141532898,
+      "step": 548
+    },
+    {
+      "epoch": 2.191235059760956,
+      "grad_norm": 0.13514572381973267,
+      "learning_rate": 2.5906638610313336e-06,
+      "loss": 1.1111763715744019,
+      "step": 550
+    },
+    {
+      "epoch": 2.199203187250996,
+      "grad_norm": 0.22583214938640594,
+      "learning_rate": 2.575870291659946e-06,
+      "loss": 0.7768072485923767,
+      "step": 552
+    },
+    {
+      "epoch": 2.2071713147410357,
+      "grad_norm": 0.2895185053348541,
+      "learning_rate": 2.5610842914058446e-06,
+      "loss": 0.4450429677963257,
+      "step": 554
+    },
+    {
+      "epoch": 2.2151394422310755,
+      "grad_norm": 0.13115426898002625,
+      "learning_rate": 2.5463065029908906e-06,
+      "loss": 0.7130606770515442,
+      "step": 556
+    },
+    {
+      "epoch": 2.2231075697211153,
+      "grad_norm": 0.27532240748405457,
+      "learning_rate": 2.5315375687799896e-06,
+      "loss": 0.8779282569885254,
+      "step": 558
+    },
+    {
+      "epoch": 2.231075697211155,
+      "grad_norm": 0.14353600144386292,
+      "learning_rate": 2.5167781307531714e-06,
+      "loss": 0.9679704308509827,
+      "step": 560
+    },
+    {
+      "epoch": 2.239043824701195,
+      "grad_norm": 0.15896089375019073,
+      "learning_rate": 2.5020288304776822e-06,
+      "loss": 1.0966684818267822,
+      "step": 562
+    },
+    {
+      "epoch": 2.247011952191235,
+      "grad_norm": 0.24172165989875793,
+      "learning_rate": 2.487290309080097e-06,
+      "loss": 0.8800580501556396,
+      "step": 564
+    },
+    {
+      "epoch": 2.2549800796812747,
+      "grad_norm": 0.13835322856903076,
+      "learning_rate": 2.472563207218454e-06,
+      "loss": 0.7232094407081604,
+      "step": 566
+    },
+    {
+      "epoch": 2.2629482071713145,
+      "grad_norm": 0.13562987744808197,
+      "learning_rate": 2.457848165054402e-06,
+      "loss": 0.7932554483413696,
+      "step": 568
+    },
+    {
+      "epoch": 2.2709163346613543,
+      "grad_norm": 0.10907457768917084,
+      "learning_rate": 2.4431458222253755e-06,
+      "loss": 1.0848677158355713,
+      "step": 570
+    },
+    {
+      "epoch": 2.278884462151394,
+      "grad_norm": 0.34634578227996826,
+      "learning_rate": 2.4284568178167923e-06,
+      "loss": 0.6181778907775879,
+      "step": 572
+    },
+    {
+      "epoch": 2.2868525896414345,
+      "grad_norm": 0.11129295080900192,
+      "learning_rate": 2.413781790334272e-06,
+      "loss": 0.7601137757301331,
+      "step": 574
+    },
+    {
+      "epoch": 2.2948207171314743,
+      "grad_norm": 0.10735038667917252,
+      "learning_rate": 2.3991213776758776e-06,
+      "loss": 0.2580484449863434,
+      "step": 576
+    },
+    {
+      "epoch": 2.302788844621514,
+      "grad_norm": 0.21603108942508698,
+      "learning_rate": 2.3844762171043975e-06,
+      "loss": 0.6623411774635315,
+      "step": 578
+    },
+    {
+      "epoch": 2.310756972111554,
+      "grad_norm": 0.1156550794839859,
+      "learning_rate": 2.3698469452196328e-06,
+      "loss": 0.6525945663452148,
+      "step": 580
+    },
+    {
+      "epoch": 2.318725099601594,
+      "grad_norm": 0.7334539890289307,
+      "learning_rate": 2.355234197930735e-06,
+      "loss": 0.6780551075935364,
+      "step": 582
+    },
+    {
+      "epoch": 2.3266932270916336,
+      "grad_norm": 0.1303965002298355,
+      "learning_rate": 2.340638610428554e-06,
+      "loss": 0.8268243074417114,
+      "step": 584
+    },
+    {
+      "epoch": 2.3346613545816735,
+      "grad_norm": 0.20556394755840302,
+      "learning_rate": 2.3260608171580412e-06,
+      "loss": 0.6557456254959106,
+      "step": 586
+    },
+    {
+      "epoch": 2.3426294820717133,
+      "grad_norm": 0.2294742912054062,
+      "learning_rate": 2.311501451790658e-06,
+      "loss": 0.9165735244750977,
+      "step": 588
+    },
+    {
+      "epoch": 2.350597609561753,
+      "grad_norm": 0.20574308931827545,
+      "learning_rate": 2.2969611471968386e-06,
+      "loss": 1.0626658201217651,
+      "step": 590
+    },
+    {
+      "epoch": 2.358565737051793,
+      "grad_norm": 0.12017904222011566,
+      "learning_rate": 2.2824405354184776e-06,
+      "loss": 1.1066505908966064,
+      "step": 592
+    },
+    {
+      "epoch": 2.366533864541833,
+      "grad_norm": 0.1420006901025772,
+      "learning_rate": 2.267940247641458e-06,
+      "loss": 0.7640745639801025,
+      "step": 594
+    },
+    {
+      "epoch": 2.3745019920318726,
+      "grad_norm": 0.1035836860537529,
+      "learning_rate": 2.2534609141682135e-06,
+      "loss": 1.0281938314437866,
+      "step": 596
+    },
+    {
+      "epoch": 2.3824701195219125,
+      "grad_norm": 0.10310082882642746,
+      "learning_rate": 2.2390031643903286e-06,
+      "loss": 0.7696731686592102,
+      "step": 598
+    },
+    {
+      "epoch": 2.3904382470119523,
+      "grad_norm": 0.3817959725856781,
+      "learning_rate": 2.224567626761185e-06,
+      "loss": 0.9735798835754395,
+      "step": 600
+    },
+    {
+      "epoch": 2.398406374501992,
+      "grad_norm": 0.11071410775184631,
+      "learning_rate": 2.2101549287686384e-06,
+      "loss": 1.0909497737884521,
+      "step": 602
+    },
+    {
+      "epoch": 2.406374501992032,
+      "grad_norm": 0.09129130095243454,
+      "learning_rate": 2.195765696907745e-06,
+      "loss": 0.6162676215171814,
+      "step": 604
+    },
+    {
+      "epoch": 2.414342629482072,
+      "grad_norm": 0.36023786664009094,
+      "learning_rate": 2.1814005566535303e-06,
+      "loss": 0.22913479804992676,
+      "step": 606
+    },
+    {
+      "epoch": 2.4223107569721116,
+      "grad_norm": 0.06375167518854141,
+      "learning_rate": 2.167060132433798e-06,
+      "loss": 0.5836560726165771,
+      "step": 608
+    },
+    {
+      "epoch": 2.4302788844621515,
+      "grad_norm": 0.06974677741527557,
+      "learning_rate": 2.1527450476019895e-06,
+      "loss": 1.0072602033615112,
+      "step": 610
+    },
+    {
+      "epoch": 2.4382470119521913,
+      "grad_norm": 0.09096641838550568,
+      "learning_rate": 2.138455924410085e-06,
+      "loss": 0.7593685388565063,
+      "step": 612
+    },
+    {
+      "epoch": 2.446215139442231,
+      "grad_norm": 0.2442321479320526,
+      "learning_rate": 2.1241933839815606e-06,
+      "loss": 0.6161373853683472,
+      "step": 614
+    },
+    {
+      "epoch": 2.454183266932271,
+      "grad_norm": 0.14059153199195862,
+      "learning_rate": 2.109958046284382e-06,
+      "loss": 0.9858022928237915,
+      "step": 616
+    },
+    {
+      "epoch": 2.462151394422311,
+      "grad_norm": 0.24893894791603088,
+      "learning_rate": 2.0957505301040604e-06,
+      "loss": 0.7891203761100769,
+      "step": 618
+    },
+    {
+      "epoch": 2.4701195219123506,
+      "grad_norm": 0.3376093804836273,
+      "learning_rate": 2.081571453016756e-06,
+      "loss": 0.8553047776222229,
+      "step": 620
+    },
+    {
+      "epoch": 2.4780876494023905,
+      "grad_norm": 0.16507503390312195,
+      "learning_rate": 2.067421431362427e-06,
+      "loss": 0.8848202228546143,
+      "step": 622
+    },
+    {
+      "epoch": 2.4860557768924303,
+      "grad_norm": 0.26530686020851135,
+      "learning_rate": 2.053301080218044e-06,
+      "loss": 0.8302243947982788,
+      "step": 624
+    },
+    {
+      "epoch": 2.49402390438247,
+      "grad_norm": 0.14043979346752167,
+      "learning_rate": 2.039211013370853e-06,
+      "loss": 0.7391160726547241,
+      "step": 626
+    },
+    {
+      "epoch": 2.50199203187251,
+      "grad_norm": 0.3858403265476227,
+      "learning_rate": 2.025151843291693e-06,
+      "loss": 1.0676630735397339,
+      "step": 628
+    },
+    {
+      "epoch": 2.50996015936255,
+      "grad_norm": 0.18159031867980957,
+      "learning_rate": 2.0111241811083733e-06,
+      "loss": 0.5425498485565186,
+      "step": 630
+    },
+    {
+      "epoch": 2.5179282868525896,
+      "grad_norm": 0.5483209490776062,
+      "learning_rate": 1.9971286365791092e-06,
+      "loss": 0.4919929504394531,
+      "step": 632
+    },
+    {
+      "epoch": 2.5258964143426295,
+      "grad_norm": 0.08576977252960205,
+      "learning_rate": 1.9831658180660197e-06,
+      "loss": 1.0784759521484375,
+      "step": 634
+    },
+    {
+      "epoch": 2.5338645418326693,
+      "grad_norm": 0.1708947718143463,
+      "learning_rate": 1.9692363325086774e-06,
+      "loss": 1.016153335571289,
+      "step": 636
+    },
+    {
+      "epoch": 2.541832669322709,
+      "grad_norm": 0.2902320325374603,
+      "learning_rate": 1.95534078539773e-06,
+      "loss": 1.0131158828735352,
+      "step": 638
+    },
+    {
+      "epoch": 2.549800796812749,
+      "grad_norm": 0.07922971248626709,
+      "learning_rate": 1.9414797807485796e-06,
+      "loss": 0.8107810020446777,
+      "step": 640
+    },
+    {
+      "epoch": 2.557768924302789,
+      "grad_norm": 0.14320027828216553,
+      "learning_rate": 1.9276539210751295e-06,
+      "loss": 1.048234224319458,
+      "step": 642
+    },
+    {
+      "epoch": 2.5657370517928286,
+      "grad_norm": 0.08111540973186493,
+      "learning_rate": 1.9138638073635886e-06,
+      "loss": 1.2002503871917725,
+      "step": 644
+    },
+    {
+      "epoch": 2.5737051792828685,
+      "grad_norm": 0.24206294119358063,
+      "learning_rate": 1.90011003904635e-06,
+      "loss": 0.5970178246498108,
+      "step": 646
+    },
+    {
+      "epoch": 2.5816733067729083,
+      "grad_norm": 0.21559061110019684,
+      "learning_rate": 1.886393213975939e-06,
+      "loss": 1.090952754020691,
+      "step": 648
+    },
+    {
+      "epoch": 2.589641434262948,
+      "grad_norm": 0.16407696902751923,
+      "learning_rate": 1.8727139283990203e-06,
+      "loss": 0.8398949503898621,
+      "step": 650
+    },
+    {
+      "epoch": 2.597609561752988,
+      "grad_norm": 0.33039212226867676,
+      "learning_rate": 1.8590727769304789e-06,
+      "loss": 0.8735826015472412,
+      "step": 652
+    },
+    {
+      "epoch": 2.605577689243028,
+      "grad_norm": 0.13958394527435303,
+      "learning_rate": 1.8454703525275814e-06,
+      "loss": 0.6662681102752686,
+      "step": 654
+    },
+    {
+      "epoch": 2.6135458167330676,
+      "grad_norm": 0.2719433605670929,
+      "learning_rate": 1.8319072464641912e-06,
+      "loss": 0.354320228099823,
+      "step": 656
+    },
+    {
+      "epoch": 2.6215139442231075,
+      "grad_norm": 0.1399993747472763,
+      "learning_rate": 1.8183840483050741e-06,
+      "loss": 0.5853203535079956,
+      "step": 658
+    },
+    {
+      "epoch": 2.6294820717131473,
+      "grad_norm": 0.11761873215436935,
+      "learning_rate": 1.8049013458802667e-06,
+      "loss": 0.919490396976471,
+      "step": 660
+    },
+    {
+      "epoch": 2.637450199203187,
+      "grad_norm": 0.20474128425121307,
+      "learning_rate": 1.7914597252595267e-06,
+      "loss": 1.097288727760315,
+      "step": 662
+    },
+    {
+      "epoch": 2.645418326693227,
+      "grad_norm": 0.2121683955192566,
+      "learning_rate": 1.7780597707268582e-06,
+      "loss": 0.5993569493293762,
+      "step": 664
+    },
+    {
+      "epoch": 2.653386454183267,
+      "grad_norm": 0.18166586756706238,
+      "learning_rate": 1.7647020647551105e-06,
+      "loss": 0.9903996586799622,
+      "step": 666
+    },
+    {
+      "epoch": 2.6613545816733066,
+      "grad_norm": 0.0722302794456482,
+      "learning_rate": 1.7513871879806605e-06,
+      "loss": 0.5933473110198975,
+      "step": 668
+    },
+    {
+      "epoch": 2.6693227091633465,
+      "grad_norm": 0.2595450282096863,
+      "learning_rate": 1.7381157191781766e-06,
+      "loss": 0.784045398235321,
+      "step": 670
+    },
+    {
+      "epoch": 2.6772908366533863,
+      "grad_norm": 0.10942405462265015,
+      "learning_rate": 1.7248882352354571e-06,
+      "loss": 0.4276721775531769,
+      "step": 672
+    },
+    {
+      "epoch": 2.685258964143426,
+      "grad_norm": 0.1339852213859558,
+      "learning_rate": 1.7117053111283528e-06,
+      "loss": 1.0860828161239624,
+      "step": 674
+    },
+    {
+      "epoch": 2.6932270916334664,
+      "grad_norm": 0.6830044984817505,
+      "learning_rate": 1.6985675198957752e-06,
+      "loss": 0.7422299385070801,
+      "step": 676
+    },
+    {
+      "epoch": 2.7011952191235062,
+      "grad_norm": 0.2023741453886032,
+      "learning_rate": 1.6854754326147916e-06,
+      "loss": 0.6293506026268005,
+      "step": 678
+    },
+    {
+      "epoch": 2.709163346613546,
+      "grad_norm": 0.13024254143238068,
+      "learning_rate": 1.6724296183757925e-06,
+      "loss": 1.0757577419281006,
+      "step": 680
+    },
+    {
+      "epoch": 2.717131474103586,
+      "grad_norm": 0.09477359801530838,
+      "learning_rate": 1.6594306442577629e-06,
+      "loss": 1.1063220500946045,
+      "step": 682
+    },
+    {
+      "epoch": 2.7250996015936257,
+      "grad_norm": 0.2803097367286682,
+      "learning_rate": 1.6464790753036298e-06,
+      "loss": 0.7243516445159912,
+      "step": 684
+    },
+    {
+      "epoch": 2.7330677290836656,
+      "grad_norm": 0.20164035260677338,
+      "learning_rate": 1.6335754744956935e-06,
+      "loss": 0.464237242937088,
+      "step": 686
+    },
+    {
+      "epoch": 2.7410358565737054,
+      "grad_norm": 0.12742824852466583,
+      "learning_rate": 1.6207204027311685e-06,
+      "loss": 1.0914649963378906,
+      "step": 688
+    },
+    {
+      "epoch": 2.7490039840637452,
+      "grad_norm": 0.23078010976314545,
+      "learning_rate": 1.6079144187977925e-06,
+      "loss": 0.5380544066429138,
+      "step": 690
+    },
+    {
+      "epoch": 2.756972111553785,
+      "grad_norm": 0.1522957980632782,
+      "learning_rate": 1.595158079349542e-06,
+      "loss": 1.1389809846878052,
+      "step": 692
+    },
+    {
+      "epoch": 2.764940239043825,
+      "grad_norm": 0.2454521209001541,
+      "learning_rate": 1.5824519388824311e-06,
+      "loss": 0.921345591545105,
+      "step": 694
+    },
+    {
+      "epoch": 2.7729083665338647,
+      "grad_norm": 0.1827179342508316,
+      "learning_rate": 1.5697965497104117e-06,
+      "loss": 0.7584734559059143,
+      "step": 696
+    },
+    {
+      "epoch": 2.7808764940239046,
+      "grad_norm": 0.11620759218931198,
+      "learning_rate": 1.557192461941367e-06,
+      "loss": 0.9769060611724854,
+      "step": 698
+    },
+    {
+      "epoch": 2.7888446215139444,
+      "grad_norm": 0.12660770118236542,
+      "learning_rate": 1.5446402234531934e-06,
+      "loss": 1.0959820747375488,
+      "step": 700
+    },
+    {
+      "epoch": 2.7968127490039842,
+      "grad_norm": 0.18615040183067322,
+      "learning_rate": 1.5321403798699924e-06,
+      "loss": 0.6809903979301453,
+      "step": 702
+    },
+    {
+      "epoch": 2.804780876494024,
+      "grad_norm": 0.16436581313610077,
+      "learning_rate": 1.5196934745383456e-06,
+      "loss": 1.3015213012695312,
+      "step": 704
+    },
+    {
+      "epoch": 2.812749003984064,
+      "grad_norm": 0.16683776676654816,
+      "learning_rate": 1.5073000485037044e-06,
+      "loss": 1.3119351863861084,
+      "step": 706
+    },
+    {
+      "epoch": 2.8207171314741037,
+      "grad_norm": 0.18298016488552094,
+      "learning_rate": 1.4949606404868638e-06,
+      "loss": 1.1386584043502808,
+      "step": 708
+    },
+    {
+      "epoch": 2.8286852589641436,
+      "grad_norm": 0.08154182881116867,
+      "learning_rate": 1.4826757868605519e-06,
+      "loss": 0.5629467368125916,
+      "step": 710
+    },
+    {
+      "epoch": 2.8366533864541834,
+      "grad_norm": 0.12383849918842316,
+      "learning_rate": 1.4704460216261125e-06,
+      "loss": 1.1116136312484741,
+      "step": 712
+    },
+    {
+      "epoch": 2.8446215139442232,
+      "grad_norm": 0.08916490525007248,
+      "learning_rate": 1.458271876390287e-06,
+      "loss": 0.6231309175491333,
+      "step": 714
+    },
+    {
+      "epoch": 2.852589641434263,
+      "grad_norm": 0.10828275978565216,
+      "learning_rate": 1.446153880342117e-06,
+      "loss": 1.0865730047225952,
+      "step": 716
+    },
+    {
+      "epoch": 2.860557768924303,
+      "grad_norm": 0.29818472266197205,
+      "learning_rate": 1.4340925602299337e-06,
+      "loss": 0.6417689919471741,
+      "step": 718
+    },
+    {
+      "epoch": 2.8685258964143427,
+      "grad_norm": 0.10018379986286163,
+      "learning_rate": 1.4220884403384656e-06,
+      "loss": 0.6161823272705078,
+      "step": 720
+    },
+    {
+      "epoch": 2.8764940239043826,
+      "grad_norm": 0.17597460746765137,
+      "learning_rate": 1.4101420424660433e-06,
+      "loss": 0.7760418653488159,
+      "step": 722
+    },
+    {
+      "epoch": 2.8844621513944224,
+      "grad_norm": 0.086418017745018,
+      "learning_rate": 1.3982538859019219e-06,
+      "loss": 0.6306876540184021,
+      "step": 724
+    },
+    {
+      "epoch": 2.8924302788844622,
+      "grad_norm": 0.14897330105304718,
+      "learning_rate": 1.3864244874037096e-06,
+      "loss": 1.0493649244308472,
+      "step": 726
+    },
+    {
+      "epoch": 2.900398406374502,
+      "grad_norm": 0.10307996720075607,
+      "learning_rate": 1.3746543611749007e-06,
+      "loss": 0.5983636379241943,
+      "step": 728
+    },
+    {
+      "epoch": 2.908366533864542,
+      "grad_norm": 0.10402888059616089,
+      "learning_rate": 1.3629440188425285e-06,
+      "loss": 1.0042794942855835,
+      "step": 730
+    },
+    {
+      "epoch": 2.9163346613545817,
+      "grad_norm": 0.2255898416042328,
+      "learning_rate": 1.351293969434921e-06,
+      "loss": 0.40956830978393555,
+      "step": 732
+    },
+    {
+      "epoch": 2.9243027888446216,
+      "grad_norm": 0.12264905124902725,
+      "learning_rate": 1.3397047193595809e-06,
+      "loss": 1.1065568923950195,
+      "step": 734
+    },
+    {
+      "epoch": 2.9322709163346614,
+      "grad_norm": 0.4319559335708618,
+      "learning_rate": 1.3281767723811651e-06,
+      "loss": 0.43042445182800293,
+      "step": 736
+    },
+    {
+      "epoch": 2.9402390438247012,
+      "grad_norm": 1.3020806312561035,
+      "learning_rate": 1.3167106295995923e-06,
+      "loss": 0.9515847563743591,
+      "step": 738
+    },
+    {
+      "epoch": 2.948207171314741,
+      "grad_norm": 0.41467341780662537,
+      "learning_rate": 1.3053067894282624e-06,
+      "loss": 1.230602741241455,
+      "step": 740
+    },
+    {
+      "epoch": 2.956175298804781,
+      "grad_norm": 0.9622936844825745,
+      "learning_rate": 1.2939657475723844e-06,
+      "loss": 0.4225766062736511,
+      "step": 742
+    },
+    {
+      "epoch": 2.9641434262948207,
+      "grad_norm": 0.10938385874032974,
+      "learning_rate": 1.2826879970074346e-06,
+      "loss": 0.9528868198394775,
+      "step": 744
+    },
+    {
+      "epoch": 2.9721115537848606,
+      "grad_norm": 0.11735125631093979,
+      "learning_rate": 1.271474027957727e-06,
+      "loss": 0.9483767151832581,
+      "step": 746
+    },
+    {
+      "epoch": 2.9800796812749004,
+      "grad_norm": 0.1643458604812622,
+      "learning_rate": 1.2603243278751043e-06,
+      "loss": 0.5863875150680542,
+      "step": 748
+    },
+    {
+      "epoch": 2.9880478087649402,
+      "grad_norm": 0.37252160906791687,
+      "learning_rate": 1.249239381417746e-06,
+      "loss": 0.1558627337217331,
+      "step": 750
+    },
+    {
+      "epoch": 2.99601593625498,
+      "grad_norm": 0.09638120979070663,
+      "learning_rate": 1.238219670429104e-06,
+      "loss": 1.0721180438995361,
+      "step": 752
+    },
+    {
+      "epoch": 3.00398406374502,
+      "grad_norm": 0.11940814554691315,
+      "learning_rate": 1.2272656739169589e-06,
+      "loss": 0.5827401280403137,
+      "step": 754
+    },
+    {
+      "epoch": 3.0119521912350598,
+      "grad_norm": 0.11712532490491867,
+      "learning_rate": 1.2163778680325948e-06,
+      "loss": 0.6694964170455933,
+      "step": 756
+    },
+    {
+      "epoch": 3.0199203187250996,
+      "grad_norm": 0.3596522808074951,
+      "learning_rate": 1.2055567260501055e-06,
+      "loss": 0.36470746994018555,
+      "step": 758
+    },
+    {
+      "epoch": 3.0278884462151394,
+      "grad_norm": 0.11949655413627625,
+      "learning_rate": 1.1948027183458182e-06,
+      "loss": 0.7314307689666748,
+      "step": 760
+    },
+    {
+      "epoch": 3.0358565737051793,
+      "grad_norm": 0.9080104231834412,
+      "learning_rate": 1.18411631237785e-06,
+      "loss": 0.4859294593334198,
+      "step": 762
+    },
+    {
+      "epoch": 3.043824701195219,
+      "grad_norm": 0.1289195716381073,
+      "learning_rate": 1.173497972665788e-06,
+      "loss": 1.0511267185211182,
+      "step": 764
+    },
+    {
+      "epoch": 3.051792828685259,
+      "grad_norm": 0.14513424038887024,
+      "learning_rate": 1.1629481607704964e-06,
+      "loss": 0.14105428755283356,
+      "step": 766
+    },
+    {
+      "epoch": 3.0597609561752988,
+      "grad_norm": 0.12946702539920807,
+      "learning_rate": 1.1524673352740543e-06,
+      "loss": 0.6402050852775574,
+      "step": 768
+    },
+    {
+      "epoch": 3.0677290836653386,
+      "grad_norm": 0.19476629793643951,
+      "learning_rate": 1.1420559517598206e-06,
+      "loss": 0.8798675537109375,
+      "step": 770
+    },
+    {
+      "epoch": 3.0756972111553784,
+      "grad_norm": 0.16828079521656036,
+      "learning_rate": 1.13171446279263e-06,
+      "loss": 1.0457466840744019,
+      "step": 772
+    },
+    {
+      "epoch": 3.0836653386454183,
+      "grad_norm": 0.3514999449253082,
+      "learning_rate": 1.1214433178991258e-06,
+      "loss": 0.9204085469245911,
+      "step": 774
+    },
+    {
+      "epoch": 3.091633466135458,
+      "grad_norm": 0.047812096774578094,
+      "learning_rate": 1.111242963548212e-06,
+      "loss": 0.6142662763595581,
+      "step": 776
+    },
+    {
+      "epoch": 3.099601593625498,
+      "grad_norm": 0.1254761964082718,
+      "learning_rate": 1.1011138431316538e-06,
+      "loss": 0.6467838883399963,
+      "step": 778
+    },
+    {
+      "epoch": 3.1075697211155378,
+      "grad_norm": 0.14199812710285187,
+      "learning_rate": 1.0910563969447969e-06,
+      "loss": 1.0296813249588013,
+      "step": 780
+    },
+    {
+      "epoch": 3.1155378486055776,
+      "grad_norm": 0.050759002566337585,
+      "learning_rate": 1.0810710621674355e-06,
+      "loss": 0.3743121027946472,
+      "step": 782
+    },
+    {
+      "epoch": 3.1235059760956174,
+      "grad_norm": 0.12713123857975006,
+      "learning_rate": 1.0711582728448034e-06,
+      "loss": 1.086698055267334,
+      "step": 784
+    },
+    {
+      "epoch": 3.1314741035856573,
+      "grad_norm": 0.12391860783100128,
+      "learning_rate": 1.0613184598687088e-06,
+      "loss": 0.824215292930603,
+      "step": 786
+    },
+    {
+      "epoch": 3.139442231075697,
+      "grad_norm": 0.3492829501628876,
+      "learning_rate": 1.0515520509588076e-06,
+      "loss": 0.6513854265213013,
+      "step": 788
+    },
+    {
+      "epoch": 3.147410358565737,
+      "grad_norm": 0.12369805574417114,
+      "learning_rate": 1.0418594706440022e-06,
+      "loss": 0.5766803026199341,
+      "step": 790
+    },
+    {
+      "epoch": 3.1553784860557768,
+      "grad_norm": 0.6963577270507812,
+      "learning_rate": 1.0322411402439982e-06,
+      "loss": 0.8416083455085754,
+      "step": 792
+    },
+    {
+      "epoch": 3.1633466135458166,
+      "grad_norm": 0.11863502115011215,
+      "learning_rate": 1.0226974778509846e-06,
+      "loss": 0.23881764709949493,
+      "step": 794
+    },
+    {
+      "epoch": 3.1713147410358564,
+      "grad_norm": 0.13583557307720184,
+      "learning_rate": 1.0132288983114611e-06,
+      "loss": 0.6774227619171143,
+      "step": 796
+    },
+    {
+      "epoch": 3.1792828685258963,
+      "grad_norm": 0.1257099211215973,
+      "learning_rate": 1.003835813208206e-06,
+      "loss": 0.5713375806808472,
+      "step": 798
+    },
+    {
+      "epoch": 3.187250996015936,
+      "grad_norm": 0.2411492019891739,
+      "learning_rate": 9.945186308423832e-07,
+      "loss": 0.8945482969284058,
+      "step": 800
+    },
+    {
+      "epoch": 3.195219123505976,
+      "grad_norm": 0.07468363642692566,
+      "learning_rate": 9.852777562157983e-07,
+      "loss": 0.780666708946228,
+      "step": 802
+    },
+    {
+      "epoch": 3.2031872509960158,
+      "grad_norm": 0.44685792922973633,
+      "learning_rate": 9.7611359101329e-07,
+      "loss": 0.7123744487762451,
+      "step": 804
+    },
+    {
+      "epoch": 3.2111553784860556,
+      "grad_norm": 0.2831321656703949,
+      "learning_rate": 9.670265335852725e-07,
+      "loss": 1.2411046028137207,
+      "step": 806
+    },
+    {
+      "epoch": 3.2191235059760954,
+      "grad_norm": 0.6657912135124207,
+      "learning_rate": 9.580169789304152e-07,
+      "loss": 0.47306421399116516,
+      "step": 808
+    },
+    {
+      "epoch": 3.2270916334661353,
+      "grad_norm": 0.21211883425712585,
+      "learning_rate": 9.490853186784801e-07,
+      "loss": 0.7701281309127808,
+      "step": 810
+    },
+    {
+      "epoch": 3.235059760956175,
+      "grad_norm": 0.16326819360256195,
+      "learning_rate": 9.402319410732911e-07,
+      "loss": 0.3617125153541565,
+      "step": 812
+    },
+    {
+      "epoch": 3.243027888446215,
+      "grad_norm": 0.15705826878547668,
+      "learning_rate": 9.314572309558617e-07,
+      "loss": 0.9077857136726379,
+      "step": 814
+    },
+    {
+      "epoch": 3.2509960159362548,
+      "grad_norm": 0.5711535811424255,
+      "learning_rate": 9.227615697476674e-07,
+      "loss": 0.3621457815170288,
+      "step": 816
+    },
+    {
+      "epoch": 3.2589641434262946,
+      "grad_norm": 0.1768590658903122,
+      "learning_rate": 9.141453354340619e-07,
+      "loss": 1.031274437904358,
+      "step": 818
+    },
+    {
+      "epoch": 3.2669322709163344,
+      "grad_norm": 1.4622611999511719,
+      "learning_rate": 9.056089025478495e-07,
+      "loss": 0.5600138306617737,
+      "step": 820
+    },
+    {
+      "epoch": 3.2749003984063743,
+      "grad_norm": 0.3159909248352051,
+      "learning_rate": 8.971526421530067e-07,
+      "loss": 0.8404796719551086,
+      "step": 822
+    },
+    {
+      "epoch": 3.2828685258964145,
+      "grad_norm": 0.2663438618183136,
+      "learning_rate": 8.887769218285469e-07,
+      "loss": 0.6064561009407043,
+      "step": 824
+    },
+    {
+      "epoch": 3.2908366533864544,
+      "grad_norm": 0.1057499498128891,
+      "learning_rate": 8.804821056525498e-07,
+      "loss": 1.040030837059021,
+      "step": 826
+    },
+    {
+      "epoch": 3.298804780876494,
+      "grad_norm": 0.17966240644454956,
+      "learning_rate": 8.722685541863284e-07,
+      "loss": 1.017094612121582,
+      "step": 828
+    },
+    {
+      "epoch": 3.306772908366534,
+      "grad_norm": 0.18758699297904968,
+      "learning_rate": 8.641366244587612e-07,
+      "loss": 0.9863868355751038,
+      "step": 830
+    },
+    {
+      "epoch": 3.314741035856574,
+      "grad_norm": 0.46482646465301514,
+      "learning_rate": 8.560866699507703e-07,
+      "loss": 0.5423831343650818,
+      "step": 832
+    },
+    {
+      "epoch": 3.3227091633466137,
+      "grad_norm": 0.2820197641849518,
+      "learning_rate": 8.481190405799569e-07,
+      "loss": 0.33546701073646545,
+      "step": 834
+    },
+    {
+      "epoch": 3.3306772908366535,
+      "grad_norm": 0.4996945261955261,
+      "learning_rate": 8.402340826853908e-07,
+      "loss": 0.5688849687576294,
+      "step": 836
+    },
+    {
+      "epoch": 3.3386454183266934,
+      "grad_norm": 0.2198592871427536,
+      "learning_rate": 8.324321390125548e-07,
+      "loss": 0.8293375968933105,
+      "step": 838
+    },
+    {
+      "epoch": 3.346613545816733,
+      "grad_norm": 0.623791515827179,
+      "learning_rate": 8.247135486984487e-07,
+      "loss": 0.20986899733543396,
+      "step": 840
+    },
+    {
+      "epoch": 3.354581673306773,
+      "grad_norm": 0.5179438591003418,
+      "learning_rate": 8.170786472568451e-07,
+      "loss": 0.8081324696540833,
+      "step": 842
+    },
+    {
+      "epoch": 3.362549800796813,
+      "grad_norm": 0.3300423324108124,
+      "learning_rate": 8.095277665637061e-07,
+      "loss": 0.7150489687919617,
+      "step": 844
+    },
+    {
+      "epoch": 3.3705179282868527,
+      "grad_norm": 0.14019058644771576,
+      "learning_rate": 8.020612348427572e-07,
+      "loss": 1.068305492401123,
+      "step": 846
+    },
+    {
+      "epoch": 3.3784860557768925,
+      "grad_norm": 0.11815018951892853,
+      "learning_rate": 7.946793766512195e-07,
+      "loss": 1.0220612287521362,
+      "step": 848
+    },
+    {
+      "epoch": 3.3864541832669324,
+      "grad_norm": 0.2657376527786255,
+      "learning_rate": 7.873825128657034e-07,
+      "loss": 1.1374307870864868,
+      "step": 850
+    },
+    {
+      "epoch": 3.394422310756972,
+      "grad_norm": 0.10612645000219345,
+      "learning_rate": 7.801709606682573e-07,
+      "loss": 1.0170397758483887,
+      "step": 852
+    },
+    {
+      "epoch": 3.402390438247012,
+      "grad_norm": 0.21848075091838837,
+      "learning_rate": 7.730450335325852e-07,
+      "loss": 0.5826485753059387,
+      "step": 854
+    },
+    {
+      "epoch": 3.410358565737052,
+      "grad_norm": 0.13201934099197388,
+      "learning_rate": 7.660050412104153e-07,
+      "loss": 0.9895707368850708,
+      "step": 856
+    },
+    {
+      "epoch": 3.4183266932270917,
+      "grad_norm": 0.10805027186870575,
+      "learning_rate": 7.5905128971804e-07,
+      "loss": 1.1645433902740479,
+      "step": 858
+    },
+    {
+      "epoch": 3.4262948207171315,
+      "grad_norm": 0.14809615910053253,
+      "learning_rate": 7.521840813230098e-07,
+      "loss": 1.097672700881958,
+      "step": 860
+    },
+    {
+      "epoch": 3.4342629482071714,
+      "grad_norm": 0.23734726011753082,
+      "learning_rate": 7.454037145309986e-07,
+      "loss": 0.7617198824882507,
+      "step": 862
+    },
+    {
+      "epoch": 3.442231075697211,
+      "grad_norm": 0.9868756532669067,
+      "learning_rate": 7.38710484072826e-07,
+      "loss": 0.6071457862854004,
+      "step": 864
+    },
+    {
+      "epoch": 3.450199203187251,
+      "grad_norm": 0.10341990739107132,
+      "learning_rate": 7.321046808916427e-07,
+      "loss": 0.8816503286361694,
+      "step": 866
+    },
+    {
+      "epoch": 3.458167330677291,
+      "grad_norm": 0.14689171314239502,
+      "learning_rate": 7.255865921302902e-07,
+      "loss": 0.807262122631073,
+      "step": 868
+    },
+    {
+      "epoch": 3.4661354581673307,
+      "grad_norm": 0.24960161745548248,
+      "learning_rate": 7.191565011188149e-07,
+      "loss": 0.5605851411819458,
+      "step": 870
+    },
+    {
+      "epoch": 3.4741035856573705,
+      "grad_norm": 0.05596492439508438,
+      "learning_rate": 7.128146873621533e-07,
+      "loss": 0.6332404017448425,
+      "step": 872
+    },
+    {
+      "epoch": 3.4820717131474104,
+      "grad_norm": 0.0735950842499733,
+      "learning_rate": 7.065614265279819e-07,
+      "loss": 0.7830731868743896,
+      "step": 874
+    },
+    {
+      "epoch": 3.49003984063745,
+      "grad_norm": 0.09377210587263107,
+      "learning_rate": 7.00396990434734e-07,
+      "loss": 0.5867562890052795,
+      "step": 876
+    },
+    {
+      "epoch": 3.49800796812749,
+      "grad_norm": 0.2778567671775818,
+      "learning_rate": 6.943216470397872e-07,
+      "loss": 0.5745432376861572,
+      "step": 878
+    },
+    {
+      "epoch": 3.50597609561753,
+      "grad_norm": 0.050860848277807236,
+      "learning_rate": 6.883356604278116e-07,
+      "loss": 0.48317959904670715,
+      "step": 880
+    },
+    {
+      "epoch": 3.5139442231075697,
+      "grad_norm": 0.2646603286266327,
+      "learning_rate": 6.824392907992938e-07,
+      "loss": 0.7294278740882874,
+      "step": 882
+    },
+    {
+      "epoch": 3.5219123505976095,
+      "grad_norm": 0.4851630628108978,
+      "learning_rate": 6.76632794459224e-07,
+      "loss": 0.23715022206306458,
+      "step": 884
+    },
+    {
+      "epoch": 3.5298804780876494,
+      "grad_norm": 0.27067816257476807,
+      "learning_rate": 6.709164238059568e-07,
+      "loss": 0.8432376384735107,
+      "step": 886
+    },
+    {
+      "epoch": 3.537848605577689,
+      "grad_norm": 0.685829758644104,
+      "learning_rate": 6.652904273202385e-07,
+      "loss": 1.0171713829040527,
+      "step": 888
+    },
+    {
+      "epoch": 3.545816733067729,
+      "grad_norm": 0.23172220587730408,
+      "learning_rate": 6.597550495544071e-07,
+      "loss": 0.5367003083229065,
+      "step": 890
+    },
+    {
+      "epoch": 3.553784860557769,
+      "grad_norm": 0.2665621340274811,
+      "learning_rate": 6.54310531121762e-07,
+      "loss": 1.0732078552246094,
+      "step": 892
+    },
+    {
+      "epoch": 3.5617529880478087,
+      "grad_norm": 0.16711026430130005,
+      "learning_rate": 6.489571086861027e-07,
+      "loss": 1.1032334566116333,
+      "step": 894
+    },
+    {
+      "epoch": 3.5697211155378485,
+      "grad_norm": 0.07729527354240417,
+      "learning_rate": 6.436950149514448e-07,
+      "loss": 0.264865517616272,
+      "step": 896
+    },
+    {
+      "epoch": 3.5776892430278884,
+      "grad_norm": 0.09780830889940262,
+      "learning_rate": 6.385244786519036e-07,
+      "loss": 0.7025995850563049,
+      "step": 898
+    },
+    {
+      "epoch": 3.585657370517928,
+      "grad_norm": 0.11390344053506851,
+      "learning_rate": 6.334457245417491e-07,
+      "loss": 0.26211053133010864,
+      "step": 900
+    },
+    {
+      "epoch": 3.593625498007968,
+      "grad_norm": 0.14651577174663544,
+      "learning_rate": 6.284589733856403e-07,
+      "loss": 0.09000970423221588,
+      "step": 902
+    },
+    {
+      "epoch": 3.601593625498008,
+      "grad_norm": 0.21179097890853882,
+      "learning_rate": 6.235644419490253e-07,
+      "loss": 0.8318758606910706,
+      "step": 904
+    },
+    {
+      "epoch": 3.6095617529880477,
+      "grad_norm": 0.2069801241159439,
+      "learning_rate": 6.18762342988722e-07,
+      "loss": 0.7049223780632019,
+      "step": 906
+    },
+    {
+      "epoch": 3.6175298804780875,
+      "grad_norm": 0.09255077689886093,
+      "learning_rate": 6.140528852436666e-07,
+      "loss": 0.9886177778244019,
+      "step": 908
+    },
+    {
+      "epoch": 3.6254980079681274,
+      "grad_norm": 0.2616262435913086,
+      "learning_rate": 6.094362734258438e-07,
+      "loss": 0.9203110933303833,
+      "step": 910
+    },
+    {
+      "epoch": 3.633466135458167,
+      "grad_norm": 0.12985344231128693,
+      "learning_rate": 6.049127082113853e-07,
+      "loss": 0.06630435585975647,
+      "step": 912
+    },
+    {
+      "epoch": 3.641434262948207,
+      "grad_norm": 0.1315426379442215,
+      "learning_rate": 6.004823862318475e-07,
+      "loss": 0.8549132347106934,
+      "step": 914
+    },
+    {
+      "epoch": 3.649402390438247,
+      "grad_norm": 0.5139184594154358,
+      "learning_rate": 5.961455000656656e-07,
+      "loss": 0.8375824093818665,
+      "step": 916
+    },
+    {
+      "epoch": 3.6573705179282867,
+      "grad_norm": 0.5944884419441223,
+      "learning_rate": 5.919022382297808e-07,
+      "loss": 0.21131211519241333,
+      "step": 918
+    },
+    {
+      "epoch": 3.6653386454183265,
+      "grad_norm": 0.13028360903263092,
+      "learning_rate": 5.877527851714479e-07,
+      "loss": 1.043203353881836,
+      "step": 920
+    },
+    {
+      "epoch": 3.6733067729083664,
+      "grad_norm": 0.9395971298217773,
+      "learning_rate": 5.836973212602144e-07,
+      "loss": 0.17550040781497955,
+      "step": 922
+    },
+    {
+      "epoch": 3.681274900398406,
+      "grad_norm": 0.22767791152000427,
+      "learning_rate": 5.797360227800832e-07,
+      "loss": 1.0971715450286865,
+      "step": 924
+    },
+    {
+      "epoch": 3.6892430278884465,
+      "grad_norm": 0.14234167337417603,
+      "learning_rate": 5.758690619218493e-07,
+      "loss": 0.3333936631679535,
+      "step": 926
+    },
+    {
+      "epoch": 3.6972111553784863,
+      "grad_norm": 0.17382486164569855,
+      "learning_rate": 5.720966067756131e-07,
+      "loss": 0.6174853444099426,
+      "step": 928
+    },
+    {
+      "epoch": 3.705179282868526,
+      "grad_norm": 0.06263237446546555,
+      "learning_rate": 5.684188213234763e-07,
+      "loss": 0.9341188073158264,
+      "step": 930
+    },
+    {
+      "epoch": 3.713147410358566,
+      "grad_norm": 0.13734297454357147,
+      "learning_rate": 5.648358654324119e-07,
+      "loss": 1.034546136856079,
+      "step": 932
+    },
+    {
+      "epoch": 3.721115537848606,
+      "grad_norm": 0.79271399974823,
+      "learning_rate": 5.613478948473166e-07,
+      "loss": 0.5859970450401306,
+      "step": 934
+    },
+    {
+      "epoch": 3.7290836653386457,
+      "grad_norm": 0.10931346565485,
+      "learning_rate": 5.579550611842395e-07,
+      "loss": 0.9486707448959351,
+      "step": 936
+    },
+    {
+      "epoch": 3.7370517928286855,
+      "grad_norm": 0.2600320875644684,
+      "learning_rate": 5.546575119237923e-07,
+      "loss": 0.772897481918335,
+      "step": 938
+    },
+    {
+      "epoch": 3.7450199203187253,
+      "grad_norm": 0.12195423990488052,
+      "learning_rate": 5.514553904047394e-07,
+      "loss": 0.8399969339370728,
+      "step": 940
+    },
+    {
+      "epoch": 3.752988047808765,
+      "grad_norm": 0.3611621856689453,
+      "learning_rate": 5.483488358177647e-07,
+      "loss": 0.8234115242958069,
+      "step": 942
+    },
+    {
+      "epoch": 3.760956175298805,
+      "grad_norm": 0.3470070958137512,
+      "learning_rate": 5.453379831994238e-07,
+      "loss": 0.5658007860183716,
+      "step": 944
+    },
+    {
+      "epoch": 3.768924302788845,
+      "grad_norm": 0.14020468294620514,
+      "learning_rate": 5.424229634262735e-07,
+      "loss": 1.051989197731018,
+      "step": 946
+    },
+    {
+      "epoch": 3.7768924302788847,
+      "grad_norm": 0.12894685566425323,
+      "learning_rate": 5.396039032091826e-07,
+      "loss": 0.6285131573677063,
+      "step": 948
+    },
+    {
+      "epoch": 3.7848605577689245,
+      "grad_norm": 0.07853851467370987,
+      "learning_rate": 5.368809250878228e-07,
+      "loss": 1.028010606765747,
+      "step": 950
+    },
+    {
+      "epoch": 3.7928286852589643,
+      "grad_norm": 0.1317497044801712,
+      "learning_rate": 5.342541474253438e-07,
+      "loss": 1.0432000160217285,
+      "step": 952
+    },
+    {
+      "epoch": 3.800796812749004,
+      "grad_norm": 0.13469719886779785,
+      "learning_rate": 5.317236844032285e-07,
+      "loss": 0.6281566023826599,
+      "step": 954
+    },
+    {
+      "epoch": 3.808764940239044,
+      "grad_norm": 0.4255220592021942,
+      "learning_rate": 5.292896460163282e-07,
+      "loss": 0.8594502210617065,
+      "step": 956
+    },
+    {
+      "epoch": 3.816733067729084,
+      "grad_norm": 0.36215320229530334,
+      "learning_rate": 5.269521380680823e-07,
+      "loss": 0.615019679069519,
+      "step": 958
+    },
+    {
+      "epoch": 3.8247011952191237,
+      "grad_norm": 0.19129793345928192,
+      "learning_rate": 5.247112621659186e-07,
+      "loss": 0.8029608726501465,
+      "step": 960
+    },
+    {
+      "epoch": 3.8326693227091635,
+      "grad_norm": 0.10067617148160934,
+      "learning_rate": 5.225671157168378e-07,
+      "loss": 0.607478141784668,
+      "step": 962
+    },
+    {
+      "epoch": 3.8406374501992033,
+      "grad_norm": 0.23532076179981232,
+      "learning_rate": 5.205197919231773e-07,
+      "loss": 1.045916199684143,
+      "step": 964
+    },
+    {
+      "epoch": 3.848605577689243,
+      "grad_norm": 0.662710428237915,
+      "learning_rate": 5.185693797785624e-07,
+      "loss": 0.40122121572494507,
+      "step": 966
+    },
+    {
+      "epoch": 3.856573705179283,
+      "grad_norm": 0.1741374135017395,
+      "learning_rate": 5.167159640640358e-07,
+      "loss": 0.5938395261764526,
+      "step": 968
+    },
+    {
+      "epoch": 3.864541832669323,
+      "grad_norm": 0.25011903047561646,
+      "learning_rate": 5.149596253443733e-07,
+      "loss": 0.8277594447135925,
+      "step": 970
+    },
+    {
+      "epoch": 3.8725099601593627,
+      "grad_norm": 0.22050832211971283,
+      "learning_rate": 5.133004399645818e-07,
+      "loss": 0.6195142269134521,
+      "step": 972
+    },
+    {
+      "epoch": 3.8804780876494025,
+      "grad_norm": 0.0799994021654129,
+      "learning_rate": 5.117384800465806e-07,
+      "loss": 0.46944645047187805,
+      "step": 974
+    },
+    {
+      "epoch": 3.8884462151394423,
+      "grad_norm": 0.10134856402873993,
+      "learning_rate": 5.102738134860656e-07,
+      "loss": 1.0364645719528198,
+      "step": 976
+    },
+    {
+      "epoch": 3.896414342629482,
+      "grad_norm": 1.9972983598709106,
+      "learning_rate": 5.089065039495598e-07,
+      "loss": 0.5864360928535461,
+      "step": 978
+    },
+    {
+      "epoch": 3.904382470119522,
+      "grad_norm": 0.08965176343917847,
+      "learning_rate": 5.076366108716438e-07,
+      "loss": 0.4354311227798462,
+      "step": 980
+    },
+    {
+      "epoch": 3.912350597609562,
+      "grad_norm": 0.16573160886764526,
+      "learning_rate": 5.064641894523737e-07,
+      "loss": 0.5563325881958008,
+      "step": 982
+    },
+    {
+      "epoch": 3.9203187250996017,
+      "grad_norm": 0.15483276546001434,
+      "learning_rate": 5.053892906548811e-07,
+      "loss": 0.9712721109390259,
+      "step": 984
+    },
+    {
+      "epoch": 3.9282868525896415,
+      "grad_norm": 0.24609829485416412,
+      "learning_rate": 5.044119612031576e-07,
+      "loss": 0.13940772414207458,
+      "step": 986
+    },
+    {
+      "epoch": 3.9362549800796813,
+      "grad_norm": 0.4988359808921814,
+      "learning_rate": 5.035322435800248e-07,
+      "loss": 0.6316894292831421,
+      "step": 988
+    },
+    {
+      "epoch": 3.944223107569721,
+      "grad_norm": 0.06558756530284882,
+      "learning_rate": 5.027501760252857e-07,
+      "loss": 0.6399556398391724,
+      "step": 990
+    },
+    {
+      "epoch": 3.952191235059761,
+      "grad_norm": 0.21805451810359955,
+      "learning_rate": 5.02065792534065e-07,
+      "loss": 1.066853642463684,
+      "step": 992
+    },
+    {
+      "epoch": 3.960159362549801,
+      "grad_norm": 0.028410326689481735,
+      "learning_rate": 5.014791228553294e-07,
+      "loss": 0.2686651349067688,
+      "step": 994
+    },
+    {
+      "epoch": 3.9681274900398407,
+      "grad_norm": 0.983101487159729,
+      "learning_rate": 5.00990192490595e-07,
+      "loss": 0.9528433680534363,
+      "step": 996
+    },
+    {
+      "epoch": 3.9760956175298805,
+      "grad_norm": 0.08354467153549194,
+      "learning_rate": 5.005990226928196e-07,
+      "loss": 0.614206075668335,
+      "step": 998
+    },
+    {
+      "epoch": 3.9840637450199203,
+      "grad_norm": 0.2843836545944214,
+      "learning_rate": 5.003056304654774e-07,
+      "loss": 0.4793241620063782,
+      "step": 1000
+    },
+    {
+      "epoch": 3.99203187250996,
+      "grad_norm": 0.10276703536510468,
+      "learning_rate": 5.001100285618214e-07,
+      "loss": 1.0421714782714844,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.05980001017451286,
+      "learning_rate": 5.000122254843282e-07,
+      "loss": 0.36267849802970886,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0,
+      "step": 1004,
+      "total_flos": 3.954132003103703e+18,
+      "train_loss": 0.909145103641359,
+      "train_runtime": 34871.4276,
+      "train_samples_per_second": 1.727,
+      "train_steps_per_second": 0.029
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1004,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 99999,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.954132003103703e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0da449856a02039da017acb91972f85a6f2a2d217b7b0f5ff9d1f0cd14011afb
+size 5649