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 +1,70 @@
+---
+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:
+- adamw lang, muon other
+- family_to_muon_lr = {
+        "language": _fallback(getattr(training_args, "language_muon_lr", 1e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_muon_lr", 1e-5), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_muon_lr", 2e-5), merger_lr),
+    }
+
+    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: 3
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- gradient_accumulation_steps: 20
+- total_train_batch_size: 60
+- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine_with_min_lr
+- lr_scheduler_warmup_steps: 0.05
+- num_epochs: 4
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 5.5.3
+- Pytorch 2.11.0+cu128
+- Datasets 4.0.0
+- Tokenizers 0.22.2

README.md

@@ -0,0 +1,70 @@
+---
+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:
+- adamw lang, muon other
+- family_to_muon_lr = {
+        "language": _fallback(getattr(training_args, "language_muon_lr", 1e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_muon_lr", 1e-5), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_muon_lr", 2e-5), merger_lr),
+    }
+
+    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: 3
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- gradient_accumulation_steps: 20
+- total_train_batch_size: 60
+- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine_with_min_lr
+- lr_scheduler_warmup_steps: 0.05
+- num_epochs: 4
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 5.5.3
+- Pytorch 2.11.0+cu128
+- Datasets 4.0.0
+- Tokenizers 0.22.2

all_results.json

@@ -0,0 +1,9 @@
+{
+    "effective_tokens_per_sec": 7414.328680986972,
+    "epoch": 4.0,
+    "total_flos": 3.519329211045118e+18,
+    "train_loss": 0.8784391032123708,
+    "train_runtime": 9185.186,
+    "train_samples_per_second": 6.558,
+    "train_steps_per_second": 0.109
+}

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 sha256:52f6334d9f3b1e3cf0c9d383b2f93769ed27efa21d23cb1c36b79179dafff0ce
+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": 7414.328680986972,
+    "epoch": 4.0,
+    "total_flos": 3.519329211045118e+18,
+    "train_loss": 0.8784391032123708,
+    "train_runtime": 9185.186,
+    "train_samples_per_second": 6.558,
+    "train_steps_per_second": 0.109
+}

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": 0.992431104183197,
+      "learning_rate": 9.803921568627452e-08,
+      "loss": 2.0965123176574707,
+      "step": 2
+    },
+    {
+      "epoch": 0.01593625498007968,
+      "grad_norm": 0.6924453377723694,
+      "learning_rate": 2.9411764705882356e-07,
+      "loss": 2.0111780166625977,
+      "step": 4
+    },
+    {
+      "epoch": 0.02390438247011952,
+      "grad_norm": 7.716898441314697,
+      "learning_rate": 4.901960784313725e-07,
+      "loss": 2.2013895511627197,
+      "step": 6
+    },
+    {
+      "epoch": 0.03187250996015936,
+      "grad_norm": 0.6880842447280884,
+      "learning_rate": 6.862745098039217e-07,
+      "loss": 1.9172271490097046,
+      "step": 8
+    },
+    {
+      "epoch": 0.0398406374501992,
+      "grad_norm": 17.071664810180664,
+      "learning_rate": 8.823529411764707e-07,
+      "loss": 1.570255994796753,
+      "step": 10
+    },
+    {
+      "epoch": 0.04780876494023904,
+      "grad_norm": 0.8794861435890198,
+      "learning_rate": 1.0784313725490197e-06,
+      "loss": 1.8185358047485352,
+      "step": 12
+    },
+    {
+      "epoch": 0.055776892430278883,
+      "grad_norm": 1.04573655128479,
+      "learning_rate": 1.2745098039215686e-06,
+      "loss": 2.72529935836792,
+      "step": 14
+    },
+    {
+      "epoch": 0.06374501992031872,
+      "grad_norm": 3.6514744758605957,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 4.22023344039917,
+      "step": 16
+    },
+    {
+      "epoch": 0.07171314741035857,
+      "grad_norm": 1.0871148109436035,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.7511793375015259,
+      "step": 18
+    },
+    {
+      "epoch": 0.0796812749003984,
+      "grad_norm": 0.7928332686424255,
+      "learning_rate": 1.8627450980392158e-06,
+      "loss": 1.6615186929702759,
+      "step": 20
+    },
+    {
+      "epoch": 0.08764940239043825,
+      "grad_norm": 0.678809404373169,
+      "learning_rate": 2.058823529411765e-06,
+      "loss": 1.561174750328064,
+      "step": 22
+    },
+    {
+      "epoch": 0.09561752988047809,
+      "grad_norm": 3.65730619430542,
+      "learning_rate": 2.254901960784314e-06,
+      "loss": 1.4451587200164795,
+      "step": 24
+    },
+    {
+      "epoch": 0.10358565737051793,
+      "grad_norm": 0.33973804116249084,
+      "learning_rate": 2.450980392156863e-06,
+      "loss": 1.549599528312683,
+      "step": 26
+    },
+    {
+      "epoch": 0.11155378486055777,
+      "grad_norm": 0.8768259286880493,
+      "learning_rate": 2.647058823529412e-06,
+      "loss": 1.6473324298858643,
+      "step": 28
+    },
+    {
+      "epoch": 0.11952191235059761,
+      "grad_norm": 2.4079723358154297,
+      "learning_rate": 2.843137254901961e-06,
+      "loss": 1.4100428819656372,
+      "step": 30
+    },
+    {
+      "epoch": 0.12749003984063745,
+      "grad_norm": 1.017041802406311,
+      "learning_rate": 3.03921568627451e-06,
+      "loss": 1.2292826175689697,
+      "step": 32
+    },
+    {
+      "epoch": 0.13545816733067728,
+      "grad_norm": 1.2878122329711914,
+      "learning_rate": 3.2352941176470594e-06,
+      "loss": 1.3186246156692505,
+      "step": 34
+    },
+    {
+      "epoch": 0.14342629482071714,
+      "grad_norm": 0.31396621465682983,
+      "learning_rate": 3.431372549019608e-06,
+      "loss": 1.4684935808181763,
+      "step": 36
+    },
+    {
+      "epoch": 0.15139442231075698,
+      "grad_norm": 1.5430070161819458,
+      "learning_rate": 3.6274509803921573e-06,
+      "loss": 1.2468595504760742,
+      "step": 38
+    },
+    {
+      "epoch": 0.1593625498007968,
+      "grad_norm": 0.5676092505455017,
+      "learning_rate": 3.8235294117647055e-06,
+      "loss": 1.374531626701355,
+      "step": 40
+    },
+    {
+      "epoch": 0.16733067729083664,
+      "grad_norm": 1.4426591396331787,
+      "learning_rate": 4.019607843137255e-06,
+      "loss": 0.8449196815490723,
+      "step": 42
+    },
+    {
+      "epoch": 0.1752988047808765,
+      "grad_norm": 0.5641521215438843,
+      "learning_rate": 4.215686274509805e-06,
+      "loss": 1.0525023937225342,
+      "step": 44
+    },
+    {
+      "epoch": 0.18326693227091634,
+      "grad_norm": 1.3331698179244995,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 0.9695785641670227,
+      "step": 46
+    },
+    {
+      "epoch": 0.19123505976095617,
+      "grad_norm": 1.4260269403457642,
+      "learning_rate": 4.607843137254902e-06,
+      "loss": 1.2222058773040771,
+      "step": 48
+    },
+    {
+      "epoch": 0.199203187250996,
+      "grad_norm": 0.5581679344177246,
+      "learning_rate": 4.803921568627452e-06,
+      "loss": 1.3815633058547974,
+      "step": 50
+    },
+    {
+      "epoch": 0.20717131474103587,
+      "grad_norm": 0.6564000844955444,
+      "learning_rate": 5e-06,
+      "loss": 1.3270047903060913,
+      "step": 52
+    },
+    {
+      "epoch": 0.2151394422310757,
+      "grad_norm": 7.490191459655762,
+      "learning_rate": 4.999951098195543e-06,
+      "loss": 1.1519843339920044,
+      "step": 54
+    },
+    {
+      "epoch": 0.22310756972111553,
+      "grad_norm": 0.8271077275276184,
+      "learning_rate": 4.9998043949078495e-06,
+      "loss": 1.547857403755188,
+      "step": 56
+    },
+    {
+      "epoch": 0.23107569721115537,
+      "grad_norm": 0.45603328943252563,
+      "learning_rate": 4.9995598965138565e-06,
+      "loss": 1.3779100179672241,
+      "step": 58
+    },
+    {
+      "epoch": 0.23904382470119523,
+      "grad_norm": 0.48901230096817017,
+      "learning_rate": 4.9992176136414875e-06,
+      "loss": 0.7783682942390442,
+      "step": 60
+    },
+    {
+      "epoch": 0.24701195219123506,
+      "grad_norm": 0.5585318803787231,
+      "learning_rate": 4.998777561169186e-06,
+      "loss": 1.3811372518539429,
+      "step": 62
+    },
+    {
+      "epoch": 0.2549800796812749,
+      "grad_norm": 0.918829619884491,
+      "learning_rate": 4.998239758225272e-06,
+      "loss": 1.1283096075057983,
+      "step": 64
+    },
+    {
+      "epoch": 0.26294820717131473,
+      "grad_norm": 0.35144737362861633,
+      "learning_rate": 4.99760422818711e-06,
+      "loss": 1.3033908605575562,
+      "step": 66
+    },
+    {
+      "epoch": 0.27091633466135456,
+      "grad_norm": 1.7168487310409546,
+      "learning_rate": 4.9968709986800925e-06,
+      "loss": 0.8534752130508423,
+      "step": 68
+    },
+    {
+      "epoch": 0.2788844621513944,
+      "grad_norm": 0.749099612236023,
+      "learning_rate": 4.996040101576436e-06,
+      "loss": 1.551970362663269,
+      "step": 70
+    },
+    {
+      "epoch": 0.2868525896414343,
+      "grad_norm": 0.7611069679260254,
+      "learning_rate": 4.995111572993804e-06,
+      "loss": 0.810857355594635,
+      "step": 72
+    },
+    {
+      "epoch": 0.2948207171314741,
+      "grad_norm": 2.7616355419158936,
+      "learning_rate": 4.994085453293725e-06,
+      "loss": 1.1411327123641968,
+      "step": 74
+    },
+    {
+      "epoch": 0.30278884462151395,
+      "grad_norm": 0.5434392094612122,
+      "learning_rate": 4.992961787079848e-06,
+      "loss": 1.4649794101715088,
+      "step": 76
+    },
+    {
+      "epoch": 0.3107569721115538,
+      "grad_norm": 1.744216799736023,
+      "learning_rate": 4.991740623196001e-06,
+      "loss": 0.8862255811691284,
+      "step": 78
+    },
+    {
+      "epoch": 0.3187250996015936,
+      "grad_norm": 0.27387645840644836,
+      "learning_rate": 4.990422014724068e-06,
+      "loss": 0.8150431513786316,
+      "step": 80
+    },
+    {
+      "epoch": 0.32669322709163345,
+      "grad_norm": 0.357980877161026,
+      "learning_rate": 4.989006018981675e-06,
+      "loss": 1.3231916427612305,
+      "step": 82
+    },
+    {
+      "epoch": 0.3346613545816733,
+      "grad_norm": 0.9259305596351624,
+      "learning_rate": 4.987492697519712e-06,
+      "loss": 0.8546282052993774,
+      "step": 84
+    },
+    {
+      "epoch": 0.3426294820717131,
+      "grad_norm": 0.5314663052558899,
+      "learning_rate": 4.985882116119643e-06,
+      "loss": 1.3115451335906982,
+      "step": 86
+    },
+    {
+      "epoch": 0.350597609561753,
+      "grad_norm": 1.0062167644500732,
+      "learning_rate": 4.9841743447906576e-06,
+      "loss": 0.8821735382080078,
+      "step": 88
+    },
+    {
+      "epoch": 0.35856573705179284,
+      "grad_norm": 0.4556940197944641,
+      "learning_rate": 4.982369457766621e-06,
+      "loss": 1.3327692747116089,
+      "step": 90
+    },
+    {
+      "epoch": 0.3665338645418327,
+      "grad_norm": 2.095075845718384,
+      "learning_rate": 4.980467533502853e-06,
+      "loss": 1.111242413520813,
+      "step": 92
+    },
+    {
+      "epoch": 0.3745019920318725,
+      "grad_norm": 0.38833457231521606,
+      "learning_rate": 4.978468654672709e-06,
+      "loss": 1.2480578422546387,
+      "step": 94
+    },
+    {
+      "epoch": 0.38247011952191234,
+      "grad_norm": 0.7908452153205872,
+      "learning_rate": 4.976372908163999e-06,
+      "loss": 1.090553879737854,
+      "step": 96
+    },
+    {
+      "epoch": 0.3904382470119522,
+      "grad_norm": 0.5664379000663757,
+      "learning_rate": 4.974180385075195e-06,
+      "loss": 1.27034592628479,
+      "step": 98
+    },
+    {
+      "epoch": 0.398406374501992,
+      "grad_norm": 0.5254221558570862,
+      "learning_rate": 4.971891180711488e-06,
+      "loss": 0.857801616191864,
+      "step": 100
+    },
+    {
+      "epoch": 0.4063745019920319,
+      "grad_norm": 0.7779160737991333,
+      "learning_rate": 4.9695053945806295e-06,
+      "loss": 1.2815327644348145,
+      "step": 102
+    },
+    {
+      "epoch": 0.41434262948207173,
+      "grad_norm": 0.3694826662540436,
+      "learning_rate": 4.9670231303886194e-06,
+      "loss": 0.7948889136314392,
+      "step": 104
+    },
+    {
+      "epoch": 0.42231075697211157,
+      "grad_norm": 0.9546980261802673,
+      "learning_rate": 4.9644444960351865e-06,
+      "loss": 0.7200071215629578,
+      "step": 106
+    },
+    {
+      "epoch": 0.4302788844621514,
+      "grad_norm": 0.38188251852989197,
+      "learning_rate": 4.9617696036091095e-06,
+      "loss": 0.982001781463623,
+      "step": 108
+    },
+    {
+      "epoch": 0.43824701195219123,
+      "grad_norm": 0.3423604667186737,
+      "learning_rate": 4.958998569383335e-06,
+      "loss": 1.2899435758590698,
+      "step": 110
+    },
+    {
+      "epoch": 0.44621513944223107,
+      "grad_norm": 1.0854977369308472,
+      "learning_rate": 4.956131513809928e-06,
+      "loss": 0.807125449180603,
+      "step": 112
+    },
+    {
+      "epoch": 0.4541832669322709,
+      "grad_norm": 0.5110481977462769,
+      "learning_rate": 4.953168561514838e-06,
+      "loss": 1.127668857574463,
+      "step": 114
+    },
+    {
+      "epoch": 0.46215139442231074,
+      "grad_norm": 0.5186073184013367,
+      "learning_rate": 4.950109841292475e-06,
+      "loss": 1.2891759872436523,
+      "step": 116
+    },
+    {
+      "epoch": 0.4701195219123506,
+      "grad_norm": 0.3974666893482208,
+      "learning_rate": 4.946955486100119e-06,
+      "loss": 1.3544564247131348,
+      "step": 118
+    },
+    {
+      "epoch": 0.47808764940239046,
+      "grad_norm": 0.4518771469593048,
+      "learning_rate": 4.943705633052134e-06,
+      "loss": 1.2342479228973389,
+      "step": 120
+    },
+    {
+      "epoch": 0.4860557768924303,
+      "grad_norm": 0.5646083950996399,
+      "learning_rate": 4.940360423414013e-06,
+      "loss": 0.9549684524536133,
+      "step": 122
+    },
+    {
+      "epoch": 0.4940239043824701,
+      "grad_norm": 1.9822407960891724,
+      "learning_rate": 4.936920002596233e-06,
+      "loss": 1.0306670665740967,
+      "step": 124
+    },
+    {
+      "epoch": 0.50199203187251,
+      "grad_norm": 0.7945346236228943,
+      "learning_rate": 4.9333845201479375e-06,
+      "loss": 0.924270749092102,
+      "step": 126
+    },
+    {
+      "epoch": 0.5099601593625498,
+      "grad_norm": 0.39196866750717163,
+      "learning_rate": 4.9297541297504345e-06,
+      "loss": 1.2751665115356445,
+      "step": 128
+    },
+    {
+      "epoch": 0.5179282868525896,
+      "grad_norm": 0.30852508544921875,
+      "learning_rate": 4.9260289892105164e-06,
+      "loss": 0.937262773513794,
+      "step": 130
+    },
+    {
+      "epoch": 0.5258964143426295,
+      "grad_norm": 0.4379243552684784,
+      "learning_rate": 4.922209260453599e-06,
+      "loss": 0.9695570468902588,
+      "step": 132
+    },
+    {
+      "epoch": 0.5338645418326693,
+      "grad_norm": 0.41885074973106384,
+      "learning_rate": 4.9182951095166875e-06,
+      "loss": 0.9263144731521606,
+      "step": 134
+    },
+    {
+      "epoch": 0.5418326693227091,
+      "grad_norm": 0.5145621299743652,
+      "learning_rate": 4.914286706541153e-06,
+      "loss": 1.2721548080444336,
+      "step": 136
+    },
+    {
+      "epoch": 0.549800796812749,
+      "grad_norm": 0.3161237835884094,
+      "learning_rate": 4.910184225765341e-06,
+      "loss": 0.9362800717353821,
+      "step": 138
+    },
+    {
+      "epoch": 0.5577689243027888,
+      "grad_norm": 1.6188466548919678,
+      "learning_rate": 4.905987845516996e-06,
+      "loss": 1.0818490982055664,
+      "step": 140
+    },
+    {
+      "epoch": 0.5657370517928287,
+      "grad_norm": 0.5755692720413208,
+      "learning_rate": 4.90169774820551e-06,
+      "loss": 1.0088565349578857,
+      "step": 142
+    },
+    {
+      "epoch": 0.5737051792828686,
+      "grad_norm": 0.37476515769958496,
+      "learning_rate": 4.897314120313996e-06,
+      "loss": 1.2249387502670288,
+      "step": 144
+    },
+    {
+      "epoch": 0.5816733067729084,
+      "grad_norm": 0.5977800488471985,
+      "learning_rate": 4.892837152391174e-06,
+      "loss": 1.3247590065002441,
+      "step": 146
+    },
+    {
+      "epoch": 0.5896414342629482,
+      "grad_norm": 0.44905203580856323,
+      "learning_rate": 4.888267039043101e-06,
+      "loss": 0.9548290967941284,
+      "step": 148
+    },
+    {
+      "epoch": 0.5976095617529881,
+      "grad_norm": 1.1062926054000854,
+      "learning_rate": 4.883603978924698e-06,
+      "loss": 0.9406803846359253,
+      "step": 150
+    },
+    {
+      "epoch": 0.6055776892430279,
+      "grad_norm": 1.2753959894180298,
+      "learning_rate": 4.878848174731126e-06,
+      "loss": 0.926374077796936,
+      "step": 152
+    },
+    {
+      "epoch": 0.6135458167330677,
+      "grad_norm": 1.027884602546692,
+      "learning_rate": 4.873999833188967e-06,
+      "loss": 1.1080248355865479,
+      "step": 154
+    },
+    {
+      "epoch": 0.6215139442231076,
+      "grad_norm": 2.7580108642578125,
+      "learning_rate": 4.8690591650472455e-06,
+      "loss": 0.8502599000930786,
+      "step": 156
+    },
+    {
+      "epoch": 0.6294820717131474,
+      "grad_norm": 0.5545320510864258,
+      "learning_rate": 4.864026385068259e-06,
+      "loss": 1.132275104522705,
+      "step": 158
+    },
+    {
+      "epoch": 0.6374501992031872,
+      "grad_norm": 0.40574920177459717,
+      "learning_rate": 4.858901712018251e-06,
+      "loss": 1.0040314197540283,
+      "step": 160
+    },
+    {
+      "epoch": 0.6454183266932271,
+      "grad_norm": 0.24671249091625214,
+      "learning_rate": 4.8536853686578965e-06,
+      "loss": 0.7447801232337952,
+      "step": 162
+    },
+    {
+      "epoch": 0.6533864541832669,
+      "grad_norm": 0.26141852140426636,
+      "learning_rate": 4.84837758173262e-06,
+      "loss": 1.139325737953186,
+      "step": 164
+    },
+    {
+      "epoch": 0.6613545816733067,
+      "grad_norm": 0.8306750059127808,
+      "learning_rate": 4.8429785819627395e-06,
+      "loss": 0.7993532419204712,
+      "step": 166
+    },
+    {
+      "epoch": 0.6693227091633466,
+      "grad_norm": 0.39537087082862854,
+      "learning_rate": 4.83748860403344e-06,
+      "loss": 1.2300182580947876,
+      "step": 168
+    },
+    {
+      "epoch": 0.6772908366533864,
+      "grad_norm": 1.5469924211502075,
+      "learning_rate": 4.831907886584566e-06,
+      "loss": 0.8583570122718811,
+      "step": 170
+    },
+    {
+      "epoch": 0.6852589641434262,
+      "grad_norm": 0.5137859582901001,
+      "learning_rate": 4.8262366722002535e-06,
+      "loss": 0.8191929459571838,
+      "step": 172
+    },
+    {
+      "epoch": 0.6932270916334662,
+      "grad_norm": 0.43556949496269226,
+      "learning_rate": 4.820475207398386e-06,
+      "loss": 1.2248395681381226,
+      "step": 174
+    },
+    {
+      "epoch": 0.701195219123506,
+      "grad_norm": 1.8620752096176147,
+      "learning_rate": 4.8146237426198735e-06,
+      "loss": 0.7317447662353516,
+      "step": 176
+    },
+    {
+      "epoch": 0.7091633466135459,
+      "grad_norm": 1.777799367904663,
+      "learning_rate": 4.808682532217772e-06,
+      "loss": 0.8984107375144958,
+      "step": 178
+    },
+    {
+      "epoch": 0.7171314741035857,
+      "grad_norm": 1.2318086624145508,
+      "learning_rate": 4.802651834446221e-06,
+      "loss": 1.059409499168396,
+      "step": 180
+    },
+    {
+      "epoch": 0.7250996015936255,
+      "grad_norm": 0.23143084347248077,
+      "learning_rate": 4.796531911449226e-06,
+      "loss": 0.4955834150314331,
+      "step": 182
+    },
+    {
+      "epoch": 0.7330677290836654,
+      "grad_norm": 0.3211841583251953,
+      "learning_rate": 4.790323029249252e-06,
+      "loss": 1.2554868459701538,
+      "step": 184
+    },
+    {
+      "epoch": 0.7410358565737052,
+      "grad_norm": 0.32706090807914734,
+      "learning_rate": 4.784025457735675e-06,
+      "loss": 1.2438033819198608,
+      "step": 186
+    },
+    {
+      "epoch": 0.749003984063745,
+      "grad_norm": 0.5452679395675659,
+      "learning_rate": 4.777639470653035e-06,
+      "loss": 1.0850731134414673,
+      "step": 188
+    },
+    {
+      "epoch": 0.7569721115537849,
+      "grad_norm": 0.9263575077056885,
+      "learning_rate": 4.771165345589148e-06,
+      "loss": 0.8628407716751099,
+      "step": 190
+    },
+    {
+      "epoch": 0.7649402390438247,
+      "grad_norm": 0.5482353568077087,
+      "learning_rate": 4.7646033639630355e-06,
+      "loss": 1.191858172416687,
+      "step": 192
+    },
+    {
+      "epoch": 0.7729083665338645,
+      "grad_norm": 0.43080008029937744,
+      "learning_rate": 4.757953811012688e-06,
+      "loss": 0.8647624850273132,
+      "step": 194
+    },
+    {
+      "epoch": 0.7808764940239044,
+      "grad_norm": 0.3980056941509247,
+      "learning_rate": 4.7512169757826735e-06,
+      "loss": 0.5781556367874146,
+      "step": 196
+    },
+    {
+      "epoch": 0.7888446215139442,
+      "grad_norm": 1.9437649250030518,
+      "learning_rate": 4.74439315111157e-06,
+      "loss": 0.7770846486091614,
+      "step": 198
+    },
+    {
+      "epoch": 0.796812749003984,
+      "grad_norm": 0.9418972134590149,
+      "learning_rate": 4.737482633619234e-06,
+      "loss": 1.0205211639404297,
+      "step": 200
+    },
+    {
+      "epoch": 0.8047808764940239,
+      "grad_norm": 0.3022700548171997,
+      "learning_rate": 4.730485723693912e-06,
+      "loss": 0.9564134478569031,
+      "step": 202
+    },
+    {
+      "epoch": 0.8127490039840638,
+      "grad_norm": 0.5594950914382935,
+      "learning_rate": 4.723402725479177e-06,
+      "loss": 1.504888892173767,
+      "step": 204
+    },
+    {
+      "epoch": 0.8207171314741036,
+      "grad_norm": 3.144648790359497,
+      "learning_rate": 4.71623394686071e-06,
+      "loss": 1.2124322652816772,
+      "step": 206
+    },
+    {
+      "epoch": 0.8286852589641435,
+      "grad_norm": 0.5176221132278442,
+      "learning_rate": 4.708979699452923e-06,
+      "loss": 0.5836835503578186,
+      "step": 208
+    },
+    {
+      "epoch": 0.8366533864541833,
+      "grad_norm": 0.4660533368587494,
+      "learning_rate": 4.701640298585405e-06,
+      "loss": 1.2577470541000366,
+      "step": 210
+    },
+    {
+      "epoch": 0.8446215139442231,
+      "grad_norm": 0.6715433597564697,
+      "learning_rate": 4.694216063289219e-06,
+      "loss": 1.2138832807540894,
+      "step": 212
+    },
+    {
+      "epoch": 0.852589641434263,
+      "grad_norm": 0.3243195712566376,
+      "learning_rate": 4.686707316283033e-06,
+      "loss": 0.9419237971305847,
+      "step": 214
+    },
+    {
+      "epoch": 0.8605577689243028,
+      "grad_norm": 0.7969087362289429,
+      "learning_rate": 4.679114383959095e-06,
+      "loss": 1.4263865947723389,
+      "step": 216
+    },
+    {
+      "epoch": 0.8685258964143426,
+      "grad_norm": 1.469743251800537,
+      "learning_rate": 4.6714375963690415e-06,
+      "loss": 0.8269627094268799,
+      "step": 218
+    },
+    {
+      "epoch": 0.8764940239043825,
+      "grad_norm": 0.40505316853523254,
+      "learning_rate": 4.6636772872095545e-06,
+      "loss": 1.232311725616455,
+      "step": 220
+    },
+    {
+      "epoch": 0.8844621513944223,
+      "grad_norm": 0.5862241983413696,
+      "learning_rate": 4.6558337938078505e-06,
+      "loss": 1.4802137613296509,
+      "step": 222
+    },
+    {
+      "epoch": 0.8924302788844621,
+      "grad_norm": 1.0030487775802612,
+      "learning_rate": 4.647907457107025e-06,
+      "loss": 0.6953156590461731,
+      "step": 224
+    },
+    {
+      "epoch": 0.900398406374502,
+      "grad_norm": 0.3918820023536682,
+      "learning_rate": 4.639898621651226e-06,
+      "loss": 0.9746219515800476,
+      "step": 226
+    },
+    {
+      "epoch": 0.9083665338645418,
+      "grad_norm": 1.1113313436508179,
+      "learning_rate": 4.631807635570681e-06,
+      "loss": 1.0772472620010376,
+      "step": 228
+    },
+    {
+      "epoch": 0.9163346613545816,
+      "grad_norm": 0.5826574563980103,
+      "learning_rate": 4.623634850566563e-06,
+      "loss": 1.1785383224487305,
+      "step": 230
+    },
+    {
+      "epoch": 0.9243027888446215,
+      "grad_norm": 0.4486997425556183,
+      "learning_rate": 4.615380621895701e-06,
+      "loss": 1.321905255317688,
+      "step": 232
+    },
+    {
+      "epoch": 0.9322709163346613,
+      "grad_norm": 0.8705466389656067,
+      "learning_rate": 4.607045308355141e-06,
+      "loss": 0.8072372674942017,
+      "step": 234
+    },
+    {
+      "epoch": 0.9402390438247012,
+      "grad_norm": 0.32237792015075684,
+      "learning_rate": 4.598629272266549e-06,
+      "loss": 0.8129540085792542,
+      "step": 236
+    },
+    {
+      "epoch": 0.9482071713147411,
+      "grad_norm": 0.3973919153213501,
+      "learning_rate": 4.590132879460458e-06,
+      "loss": 0.42645901441574097,
+      "step": 238
+    },
+    {
+      "epoch": 0.9561752988047809,
+      "grad_norm": 6.2584686279296875,
+      "learning_rate": 4.5815564992603705e-06,
+      "loss": 0.839176595211029,
+      "step": 240
+    },
+    {
+      "epoch": 0.9641434262948207,
+      "grad_norm": 0.7081606984138489,
+      "learning_rate": 4.572900504466702e-06,
+      "loss": 1.1235440969467163,
+      "step": 242
+    },
+    {
+      "epoch": 0.9721115537848606,
+      "grad_norm": 0.39649632573127747,
+      "learning_rate": 4.564165271340577e-06,
+      "loss": 1.2115644216537476,
+      "step": 244
+    },
+    {
+      "epoch": 0.9800796812749004,
+      "grad_norm": 1.6367366313934326,
+      "learning_rate": 4.555351179587472e-06,
+      "loss": 0.9500784873962402,
+      "step": 246
+    },
+    {
+      "epoch": 0.9880478087649402,
+      "grad_norm": 0.46789565682411194,
+      "learning_rate": 4.546458612340714e-06,
+      "loss": 0.9774641990661621,
+      "step": 248
+    },
+    {
+      "epoch": 0.9960159362549801,
+      "grad_norm": 0.4159293472766876,
+      "learning_rate": 4.537487956144821e-06,
+      "loss": 0.9703260064125061,
+      "step": 250
+    },
+    {
+      "epoch": 1.00398406374502,
+      "grad_norm": 3.030120611190796,
+      "learning_rate": 4.528439600938706e-06,
+      "loss": 0.9079171419143677,
+      "step": 252
+    },
+    {
+      "epoch": 1.0119521912350598,
+      "grad_norm": 0.3234848380088806,
+      "learning_rate": 4.519313940038721e-06,
+      "loss": 0.7094411849975586,
+      "step": 254
+    },
+    {
+      "epoch": 1.0199203187250996,
+      "grad_norm": 0.5853373408317566,
+      "learning_rate": 4.510111370121563e-06,
+      "loss": 1.172021746635437,
+      "step": 256
+    },
+    {
+      "epoch": 1.0278884462151394,
+      "grad_norm": 0.511910617351532,
+      "learning_rate": 4.500832291207033e-06,
+      "loss": 1.1384598016738892,
+      "step": 258
+    },
+    {
+      "epoch": 1.0358565737051793,
+      "grad_norm": 0.7625296711921692,
+      "learning_rate": 4.491477106640646e-06,
+      "loss": 1.091207504272461,
+      "step": 260
+    },
+    {
+      "epoch": 1.043824701195219,
+      "grad_norm": 0.47513219714164734,
+      "learning_rate": 4.482046223076092e-06,
+      "loss": 1.01276433467865,
+      "step": 262
+    },
+    {
+      "epoch": 1.051792828685259,
+      "grad_norm": 2.4036881923675537,
+      "learning_rate": 4.4725400504575725e-06,
+      "loss": 1.1120725870132446,
+      "step": 264
+    },
+    {
+      "epoch": 1.0597609561752988,
+      "grad_norm": 1.344622254371643,
+      "learning_rate": 4.462959002001971e-06,
+      "loss": 0.6211552023887634,
+      "step": 266
+    },
+    {
+      "epoch": 1.0677290836653386,
+      "grad_norm": 0.3280583918094635,
+      "learning_rate": 4.453303494180895e-06,
+      "loss": 0.9181345701217651,
+      "step": 268
+    },
+    {
+      "epoch": 1.0756972111553784,
+      "grad_norm": 0.9431546330451965,
+      "learning_rate": 4.443573946702573e-06,
+      "loss": 0.5670726299285889,
+      "step": 270
+    },
+    {
+      "epoch": 1.0836653386454183,
+      "grad_norm": 0.41420289874076843,
+      "learning_rate": 4.433770782493606e-06,
+      "loss": 1.2278110980987549,
+      "step": 272
+    },
+    {
+      "epoch": 1.091633466135458,
+      "grad_norm": 1.8112084865570068,
+      "learning_rate": 4.42389442768059e-06,
+      "loss": 0.6097539067268372,
+      "step": 274
+    },
+    {
+      "epoch": 1.099601593625498,
+      "grad_norm": 0.39676377177238464,
+      "learning_rate": 4.413945311571586e-06,
+      "loss": 0.9101986885070801,
+      "step": 276
+    },
+    {
+      "epoch": 1.1075697211155378,
+      "grad_norm": 0.3688848614692688,
+      "learning_rate": 4.40392386663747e-06,
+      "loss": 0.8728987574577332,
+      "step": 278
+    },
+    {
+      "epoch": 1.1155378486055776,
+      "grad_norm": 0.4303763806819916,
+      "learning_rate": 4.393830528493118e-06,
+      "loss": 1.0260114669799805,
+      "step": 280
+    },
+    {
+      "epoch": 1.1235059760956174,
+      "grad_norm": 0.47700682282447815,
+      "learning_rate": 4.3836657358784875e-06,
+      "loss": 1.1537423133850098,
+      "step": 282
+    },
+    {
+      "epoch": 1.1314741035856573,
+      "grad_norm": 0.5435411334037781,
+      "learning_rate": 4.373429930639534e-06,
+      "loss": 1.176669955253601,
+      "step": 284
+    },
+    {
+      "epoch": 1.139442231075697,
+      "grad_norm": 0.2700730562210083,
+      "learning_rate": 4.3631235577090085e-06,
+      "loss": 0.2517479360103607,
+      "step": 286
+    },
+    {
+      "epoch": 1.1474103585657371,
+      "grad_norm": 1.0374253988265991,
+      "learning_rate": 4.35274706508712e-06,
+      "loss": 0.7267593145370483,
+      "step": 288
+    },
+    {
+      "epoch": 1.155378486055777,
+      "grad_norm": 1.0754668712615967,
+      "learning_rate": 4.342300903822057e-06,
+      "loss": 0.9300534129142761,
+      "step": 290
+    },
+    {
+      "epoch": 1.1633466135458168,
+      "grad_norm": 0.4998929500579834,
+      "learning_rate": 4.331785527990384e-06,
+      "loss": 0.4582483768463135,
+      "step": 292
+    },
+    {
+      "epoch": 1.1713147410358566,
+      "grad_norm": 0.44691503047943115,
+      "learning_rate": 4.3212013946773045e-06,
+      "loss": 1.1829777956008911,
+      "step": 294
+    },
+    {
+      "epoch": 1.1792828685258965,
+      "grad_norm": 0.36078473925590515,
+      "learning_rate": 4.31054896395679e-06,
+      "loss": 1.2210935354232788,
+      "step": 296
+    },
+    {
+      "epoch": 1.1872509960159363,
+      "grad_norm": 0.4337320625782013,
+      "learning_rate": 4.299828698871581e-06,
+      "loss": 1.339713215827942,
+      "step": 298
+    },
+    {
+      "epoch": 1.1952191235059761,
+      "grad_norm": 1.310649037361145,
+      "learning_rate": 4.289041065413062e-06,
+      "loss": 0.6005781292915344,
+      "step": 300
+    },
+    {
+      "epoch": 1.203187250996016,
+      "grad_norm": 0.7879869341850281,
+      "learning_rate": 4.278186532501003e-06,
+      "loss": 0.905928373336792,
+      "step": 302
+    },
+    {
+      "epoch": 1.2111553784860558,
+      "grad_norm": 0.995727002620697,
+      "learning_rate": 4.267265571963179e-06,
+      "loss": 0.5665251612663269,
+      "step": 304
+    },
+    {
+      "epoch": 1.2191235059760956,
+      "grad_norm": 0.6685842275619507,
+      "learning_rate": 4.256278658514858e-06,
+      "loss": 1.1679246425628662,
+      "step": 306
+    },
+    {
+      "epoch": 1.2270916334661355,
+      "grad_norm": 0.8777851462364197,
+      "learning_rate": 4.245226269738165e-06,
+      "loss": 1.0079498291015625,
+      "step": 308
+    },
+    {
+      "epoch": 1.2350597609561753,
+      "grad_norm": 1.0120943784713745,
+      "learning_rate": 4.23410888606133e-06,
+      "loss": 1.1408251523971558,
+      "step": 310
+    },
+    {
+      "epoch": 1.2430278884462151,
+      "grad_norm": 1.1217892169952393,
+      "learning_rate": 4.222926990737793e-06,
+      "loss": 1.0556583404541016,
+      "step": 312
+    },
+    {
+      "epoch": 1.250996015936255,
+      "grad_norm": 0.8425201177597046,
+      "learning_rate": 4.211681069825207e-06,
+      "loss": 0.875999927520752,
+      "step": 314
+    },
+    {
+      "epoch": 1.2589641434262948,
+      "grad_norm": 1.5495978593826294,
+      "learning_rate": 4.200371612164305e-06,
+      "loss": 0.7391699552536011,
+      "step": 316
+    },
+    {
+      "epoch": 1.2669322709163346,
+      "grad_norm": 3.013714075088501,
+      "learning_rate": 4.1889991093576555e-06,
+      "loss": 0.6309952735900879,
+      "step": 318
+    },
+    {
+      "epoch": 1.2749003984063745,
+      "grad_norm": 1.1000335216522217,
+      "learning_rate": 4.177564055748284e-06,
+      "loss": 0.5071834325790405,
+      "step": 320
+    },
+    {
+      "epoch": 1.2828685258964143,
+      "grad_norm": 1.0480402708053589,
+      "learning_rate": 4.166066948398199e-06,
+      "loss": 0.981095016002655,
+      "step": 322
+    },
+    {
+      "epoch": 1.2908366533864541,
+      "grad_norm": 1.4282066822052002,
+      "learning_rate": 4.154508287066772e-06,
+      "loss": 0.9262098073959351,
+      "step": 324
+    },
+    {
+      "epoch": 1.298804780876494,
+      "grad_norm": 0.57856285572052,
+      "learning_rate": 4.142888574189025e-06,
+      "loss": 1.2803043127059937,
+      "step": 326
+    },
+    {
+      "epoch": 1.3067729083665338,
+      "grad_norm": 0.41403651237487793,
+      "learning_rate": 4.131208314853778e-06,
+      "loss": 0.7203609943389893,
+      "step": 328
+    },
+    {
+      "epoch": 1.3147410358565736,
+      "grad_norm": 0.4426195025444031,
+      "learning_rate": 4.119468016781707e-06,
+      "loss": 0.8939486145973206,
+      "step": 330
+    },
+    {
+      "epoch": 1.3227091633466135,
+      "grad_norm": 1.1591002941131592,
+      "learning_rate": 4.107668190303266e-06,
+      "loss": 0.9234914183616638,
+      "step": 332
+    },
+    {
+      "epoch": 1.3306772908366533,
+      "grad_norm": 1.6737353801727295,
+      "learning_rate": 4.095809348336506e-06,
+      "loss": 0.6921515464782715,
+      "step": 334
+    },
+    {
+      "epoch": 1.3386454183266931,
+      "grad_norm": 1.4772138595581055,
+      "learning_rate": 4.083892006364777e-06,
+      "loss": 0.8849189281463623,
+      "step": 336
+    },
+    {
+      "epoch": 1.3466135458167332,
+      "grad_norm": 1.2881124019622803,
+      "learning_rate": 4.071916682414327e-06,
+      "loss": 1.1362943649291992,
+      "step": 338
+    },
+    {
+      "epoch": 1.354581673306773,
+      "grad_norm": 0.4266433119773865,
+      "learning_rate": 4.059883897031777e-06,
+      "loss": 1.1299281120300293,
+      "step": 340
+    },
+    {
+      "epoch": 1.3625498007968129,
+      "grad_norm": 0.4193069636821747,
+      "learning_rate": 4.0477941732615e-06,
+      "loss": 1.1752209663391113,
+      "step": 342
+    },
+    {
+      "epoch": 1.3705179282868527,
+      "grad_norm": 1.072519063949585,
+      "learning_rate": 4.035648036622879e-06,
+      "loss": 1.169982671737671,
+      "step": 344
+    },
+    {
+      "epoch": 1.3784860557768925,
+      "grad_norm": 1.1516021490097046,
+      "learning_rate": 4.023446015087469e-06,
+      "loss": 0.7040839195251465,
+      "step": 346
+    },
+    {
+      "epoch": 1.3864541832669324,
+      "grad_norm": 0.5228333473205566,
+      "learning_rate": 4.0111886390560446e-06,
+      "loss": 1.163697600364685,
+      "step": 348
+    },
+    {
+      "epoch": 1.3944223107569722,
+      "grad_norm": 0.2819441854953766,
+      "learning_rate": 3.998876441335544e-06,
+      "loss": 0.6836432814598083,
+      "step": 350
+    },
+    {
+      "epoch": 1.402390438247012,
+      "grad_norm": 0.36361777782440186,
+      "learning_rate": 3.98650995711591e-06,
+      "loss": 1.1881886720657349,
+      "step": 352
+    },
+    {
+      "epoch": 1.4103585657370519,
+      "grad_norm": 0.5527525544166565,
+      "learning_rate": 3.974089723946826e-06,
+      "loss": 1.1089881658554077,
+      "step": 354
+    },
+    {
+      "epoch": 1.4183266932270917,
+      "grad_norm": 0.9690864086151123,
+      "learning_rate": 3.9616162817143465e-06,
+      "loss": 0.9128237962722778,
+      "step": 356
+    },
+    {
+      "epoch": 1.4262948207171315,
+      "grad_norm": 0.7041119933128357,
+      "learning_rate": 3.949090172617437e-06,
+      "loss": 1.1476466655731201,
+      "step": 358
+    },
+    {
+      "epoch": 1.4342629482071714,
+      "grad_norm": 0.5612359046936035,
+      "learning_rate": 3.936511941144397e-06,
+      "loss": 0.6003270745277405,
+      "step": 360
+    },
+    {
+      "epoch": 1.4422310756972112,
+      "grad_norm": 0.7487348318099976,
+      "learning_rate": 3.9238821340491936e-06,
+      "loss": 0.7163197994232178,
+      "step": 362
+    },
+    {
+      "epoch": 1.450199203187251,
+      "grad_norm": 0.9313085675239563,
+      "learning_rate": 3.911201300327703e-06,
+      "loss": 1.0451388359069824,
+      "step": 364
+    },
+    {
+      "epoch": 1.4581673306772909,
+      "grad_norm": 0.6188122630119324,
+      "learning_rate": 3.898469991193835e-06,
+      "loss": 1.1301738023757935,
+      "step": 366
+    },
+    {
+      "epoch": 1.4661354581673307,
+      "grad_norm": 0.40409159660339355,
+      "learning_rate": 3.885688760055581e-06,
+      "loss": 0.6099163889884949,
+      "step": 368
+    },
+    {
+      "epoch": 1.4741035856573705,
+      "grad_norm": 0.3847958743572235,
+      "learning_rate": 3.872858162490956e-06,
+      "loss": 0.671179473400116,
+      "step": 370
+    },
+    {
+      "epoch": 1.4820717131474104,
+      "grad_norm": 0.5476490259170532,
+      "learning_rate": 3.859978756223845e-06,
+      "loss": 1.220924735069275,
+      "step": 372
+    },
+    {
+      "epoch": 1.4900398406374502,
+      "grad_norm": 0.47861382365226746,
+      "learning_rate": 3.847051101099768e-06,
+      "loss": 1.144132375717163,
+      "step": 374
+    },
+    {
+      "epoch": 1.49800796812749,
+      "grad_norm": 0.37773212790489197,
+      "learning_rate": 3.834075759061535e-06,
+      "loss": 1.1360963582992554,
+      "step": 376
+    },
+    {
+      "epoch": 1.5059760956175299,
+      "grad_norm": 1.0848755836486816,
+      "learning_rate": 3.8210532941248255e-06,
+      "loss": 0.9269629716873169,
+      "step": 378
+    },
+    {
+      "epoch": 1.5139442231075697,
+      "grad_norm": 0.46528637409210205,
+      "learning_rate": 3.8079842723536688e-06,
+      "loss": 1.1278562545776367,
+      "step": 380
+    },
+    {
+      "epoch": 1.5219123505976095,
+      "grad_norm": 0.4154779613018036,
+      "learning_rate": 3.7948692618358424e-06,
+      "loss": 0.7394702434539795,
+      "step": 382
+    },
+    {
+      "epoch": 1.5298804780876494,
+      "grad_norm": 0.3956635594367981,
+      "learning_rate": 3.781708832658169e-06,
+      "loss": 0.7812275886535645,
+      "step": 384
+    },
+    {
+      "epoch": 1.5378486055776892,
+      "grad_norm": 1.2779585123062134,
+      "learning_rate": 3.7685035568817492e-06,
+      "loss": 1.0294419527053833,
+      "step": 386
+    },
+    {
+      "epoch": 1.545816733067729,
+      "grad_norm": 0.6137291789054871,
+      "learning_rate": 3.7552540085170826e-06,
+      "loss": 1.146806240081787,
+      "step": 388
+    },
+    {
+      "epoch": 1.5537848605577689,
+      "grad_norm": 1.1068216562271118,
+      "learning_rate": 3.7419607634991235e-06,
+      "loss": 0.5710906982421875,
+      "step": 390
+    },
+    {
+      "epoch": 1.5617529880478087,
+      "grad_norm": 1.3095977306365967,
+      "learning_rate": 3.7286243996622432e-06,
+      "loss": 0.5845750570297241,
+      "step": 392
+    },
+    {
+      "epoch": 1.5697211155378485,
+      "grad_norm": 1.5929396152496338,
+      "learning_rate": 3.715245496715116e-06,
+      "loss": 1.115755319595337,
+      "step": 394
+    },
+    {
+      "epoch": 1.5776892430278884,
+      "grad_norm": 1.1586883068084717,
+      "learning_rate": 3.7018246362155134e-06,
+      "loss": 1.0371090173721313,
+      "step": 396
+    },
+    {
+      "epoch": 1.5856573705179282,
+      "grad_norm": 0.42481380701065063,
+      "learning_rate": 3.6883624015450344e-06,
+      "loss": 1.1346615552902222,
+      "step": 398
+    },
+    {
+      "epoch": 1.593625498007968,
+      "grad_norm": 0.5337651968002319,
+      "learning_rate": 3.6748593778837367e-06,
+      "loss": 0.6428213715553284,
+      "step": 400
+    },
+    {
+      "epoch": 1.6015936254980079,
+      "grad_norm": 0.7591474652290344,
+      "learning_rate": 3.661316152184708e-06,
+      "loss": 1.1169513463974,
+      "step": 402
+    },
+    {
+      "epoch": 1.6095617529880477,
+      "grad_norm": 0.6696905493736267,
+      "learning_rate": 3.6477333131485483e-06,
+      "loss": 1.2621333599090576,
+      "step": 404
+    },
+    {
+      "epoch": 1.6175298804780875,
+      "grad_norm": 1.6090987920761108,
+      "learning_rate": 3.634111451197781e-06,
+      "loss": 0.7362015247344971,
+      "step": 406
+    },
+    {
+      "epoch": 1.6254980079681274,
+      "grad_norm": 0.3145546615123749,
+      "learning_rate": 3.6204511584511865e-06,
+      "loss": 0.5767532587051392,
+      "step": 408
+    },
+    {
+      "epoch": 1.6334661354581672,
+      "grad_norm": 0.2829693853855133,
+      "learning_rate": 3.606753028698067e-06,
+      "loss": 0.7675127983093262,
+      "step": 410
+    },
+    {
+      "epoch": 1.641434262948207,
+      "grad_norm": 0.48404109477996826,
+      "learning_rate": 3.593017657372436e-06,
+      "loss": 0.4761987328529358,
+      "step": 412
+    },
+    {
+      "epoch": 1.6494023904382469,
+      "grad_norm": 0.3886066973209381,
+      "learning_rate": 3.5792456415271288e-06,
+      "loss": 0.8401291966438293,
+      "step": 414
+    },
+    {
+      "epoch": 1.6573705179282867,
+      "grad_norm": 1.244385004043579,
+      "learning_rate": 3.565437579807858e-06,
+      "loss": 0.7371784448623657,
+      "step": 416
+    },
+    {
+      "epoch": 1.6653386454183265,
+      "grad_norm": 0.7426787614822388,
+      "learning_rate": 3.551594072427188e-06,
+      "loss": 0.9862039089202881,
+      "step": 418
+    },
+    {
+      "epoch": 1.6733067729083664,
+      "grad_norm": 2.1551995277404785,
+      "learning_rate": 3.537715721138444e-06,
+      "loss": 1.138071894645691,
+      "step": 420
+    },
+    {
+      "epoch": 1.6812749003984062,
+      "grad_norm": 0.42246100306510925,
+      "learning_rate": 3.523803129209554e-06,
+      "loss": 1.0254541635513306,
+      "step": 422
+    },
+    {
+      "epoch": 1.6892430278884463,
+      "grad_norm": 0.5328935384750366,
+      "learning_rate": 3.5098569013968314e-06,
+      "loss": 0.5915545225143433,
+      "step": 424
+    },
+    {
+      "epoch": 1.697211155378486,
+      "grad_norm": 0.5242877006530762,
+      "learning_rate": 3.49587764391868e-06,
+      "loss": 1.2012101411819458,
+      "step": 426
+    },
+    {
+      "epoch": 1.705179282868526,
+      "grad_norm": 0.4879375100135803,
+      "learning_rate": 3.4818659644292473e-06,
+      "loss": 1.135599970817566,
+      "step": 428
+    },
+    {
+      "epoch": 1.7131474103585658,
+      "grad_norm": 0.6228365302085876,
+      "learning_rate": 3.4678224719920085e-06,
+      "loss": 0.7497591376304626,
+      "step": 430
+    },
+    {
+      "epoch": 1.7211155378486056,
+      "grad_norm": 0.6589022874832153,
+      "learning_rate": 3.4537477770532933e-06,
+      "loss": 0.9566125273704529,
+      "step": 432
+    },
+    {
+      "epoch": 1.7290836653386454,
+      "grad_norm": 1.0548697710037231,
+      "learning_rate": 3.4396424914157527e-06,
+      "loss": 0.6920331716537476,
+      "step": 434
+    },
+    {
+      "epoch": 1.7370517928286853,
+      "grad_norm": 1.4375154972076416,
+      "learning_rate": 3.4255072282117574e-06,
+      "loss": 0.363805890083313,
+      "step": 436
+    },
+    {
+      "epoch": 1.745019920318725,
+      "grad_norm": 0.5286189913749695,
+      "learning_rate": 3.411342601876754e-06,
+      "loss": 0.7935215830802917,
+      "step": 438
+    },
+    {
+      "epoch": 1.752988047808765,
+      "grad_norm": 4.52565336227417,
+      "learning_rate": 3.3971492281225554e-06,
+      "loss": 0.729618489742279,
+      "step": 440
+    },
+    {
+      "epoch": 1.7609561752988048,
+      "grad_norm": 0.696873128414154,
+      "learning_rate": 3.382927723910572e-06,
+      "loss": 1.4971410036087036,
+      "step": 442
+    },
+    {
+      "epoch": 1.7689243027888446,
+      "grad_norm": 0.5262324810028076,
+      "learning_rate": 3.3686787074249966e-06,
+      "loss": 0.7912355065345764,
+      "step": 444
+    },
+    {
+      "epoch": 1.7768924302788844,
+      "grad_norm": 0.3645341396331787,
+      "learning_rate": 3.354402798045935e-06,
+      "loss": 1.0550092458724976,
+      "step": 446
+    },
+    {
+      "epoch": 1.7848605577689243,
+      "grad_norm": 0.8887290954589844,
+      "learning_rate": 3.340100616322479e-06,
+      "loss": 1.1042665243148804,
+      "step": 448
+    },
+    {
+      "epoch": 1.792828685258964,
+      "grad_norm": 0.954005241394043,
+      "learning_rate": 3.3257727839457356e-06,
+      "loss": 0.8549948930740356,
+      "step": 450
+    },
+    {
+      "epoch": 1.800796812749004,
+      "grad_norm": 0.4765973389148712,
+      "learning_rate": 3.311419923721798e-06,
+      "loss": 0.5229886770248413,
+      "step": 452
+    },
+    {
+      "epoch": 1.8087649402390438,
+      "grad_norm": 15.211450576782227,
+      "learning_rate": 3.29704265954468e-06,
+      "loss": 0.9388401508331299,
+      "step": 454
+    },
+    {
+      "epoch": 1.8167330677290838,
+      "grad_norm": 0.5665377974510193,
+      "learning_rate": 3.282641616369193e-06,
+      "loss": 0.6281580924987793,
+      "step": 456
+    },
+    {
+      "epoch": 1.8247011952191237,
+      "grad_norm": 0.5543078184127808,
+      "learning_rate": 3.2682174201837773e-06,
+      "loss": 1.0435411930084229,
+      "step": 458
+    },
+    {
+      "epoch": 1.8326693227091635,
+      "grad_norm": 0.7510952949523926,
+      "learning_rate": 3.2537706979832955e-06,
+      "loss": 0.7342970371246338,
+      "step": 460
+    },
+    {
+      "epoch": 1.8406374501992033,
+      "grad_norm": 0.5338124632835388,
+      "learning_rate": 3.23930207774178e-06,
+      "loss": 0.7850310206413269,
+      "step": 462
+    },
+    {
+      "epoch": 1.8486055776892432,
+      "grad_norm": 0.25599774718284607,
+      "learning_rate": 3.2248121883851296e-06,
+      "loss": 0.11079226434230804,
+      "step": 464
+    },
+    {
+      "epoch": 1.856573705179283,
+      "grad_norm": 0.7927799820899963,
+      "learning_rate": 3.210301659763777e-06,
+      "loss": 1.156887173652649,
+      "step": 466
+    },
+    {
+      "epoch": 1.8645418326693228,
+      "grad_norm": 2.078766345977783,
+      "learning_rate": 3.1957711226253073e-06,
+      "loss": 0.5489357113838196,
+      "step": 468
+    },
+    {
+      "epoch": 1.8725099601593627,
+      "grad_norm": 0.6955522894859314,
+      "learning_rate": 3.181221208587041e-06,
+      "loss": 1.1430308818817139,
+      "step": 470
+    },
+    {
+      "epoch": 1.8804780876494025,
+      "grad_norm": 1.3641225099563599,
+      "learning_rate": 3.16665255010858e-06,
+      "loss": 1.0507137775421143,
+      "step": 472
+    },
+    {
+      "epoch": 1.8884462151394423,
+      "grad_norm": 0.9133515954017639,
+      "learning_rate": 3.152065780464311e-06,
+      "loss": 1.1702569723129272,
+      "step": 474
+    },
+    {
+      "epoch": 1.8964143426294822,
+      "grad_norm": 0.29633691906929016,
+      "learning_rate": 3.1374615337158874e-06,
+      "loss": 0.9763537645339966,
+      "step": 476
+    },
+    {
+      "epoch": 1.904382470119522,
+      "grad_norm": 0.428254634141922,
+      "learning_rate": 3.1228404446846577e-06,
+      "loss": 0.8701446056365967,
+      "step": 478
+    },
+    {
+      "epoch": 1.9123505976095618,
+      "grad_norm": 1.131527304649353,
+      "learning_rate": 3.1082031489240775e-06,
+      "loss": 0.611979603767395,
+      "step": 480
+    },
+    {
+      "epoch": 1.9203187250996017,
+      "grad_norm": 0.7094573378562927,
+      "learning_rate": 3.093550282692081e-06,
+      "loss": 0.7528533339500427,
+      "step": 482
+    },
+    {
+      "epoch": 1.9282868525896415,
+      "grad_norm": 1.105019450187683,
+      "learning_rate": 3.0788824829234206e-06,
+      "loss": 1.1147563457489014,
+      "step": 484
+    },
+    {
+      "epoch": 1.9362549800796813,
+      "grad_norm": 0.5338658094406128,
+      "learning_rate": 3.064200387201987e-06,
+      "loss": 1.148236870765686,
+      "step": 486
+    },
+    {
+      "epoch": 1.9442231075697212,
+      "grad_norm": 0.5745646953582764,
+      "learning_rate": 3.0495046337330917e-06,
+      "loss": 1.1073964834213257,
+      "step": 488
+    },
+    {
+      "epoch": 1.952191235059761,
+      "grad_norm": 0.3760293424129486,
+      "learning_rate": 3.034795861315723e-06,
+      "loss": 1.117495059967041,
+      "step": 490
+    },
+    {
+      "epoch": 1.9601593625498008,
+      "grad_norm": 0.6365296840667725,
+      "learning_rate": 3.020074709314781e-06,
+      "loss": 0.7863898873329163,
+      "step": 492
+    },
+    {
+      "epoch": 1.9681274900398407,
+      "grad_norm": 2.4303553104400635,
+      "learning_rate": 3.0053418176332844e-06,
+      "loss": 0.6176267266273499,
+      "step": 494
+    },
+    {
+      "epoch": 1.9760956175298805,
+      "grad_norm": 0.3201427161693573,
+      "learning_rate": 2.9905978266845565e-06,
+      "loss": 1.0759797096252441,
+      "step": 496
+    },
+    {
+      "epoch": 1.9840637450199203,
+      "grad_norm": 1.283196210861206,
+      "learning_rate": 2.975843377364385e-06,
+      "loss": 0.8235762715339661,
+      "step": 498
+    },
+    {
+      "epoch": 1.9920318725099602,
+      "grad_norm": 0.8943160772323608,
+      "learning_rate": 2.961079111023167e-06,
+      "loss": 1.1678543090820312,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6844873428344727,
+      "learning_rate": 2.9463056694380263e-06,
+      "loss": 0.6115949749946594,
+      "step": 502
+    },
+    {
+      "epoch": 2.00796812749004,
+      "grad_norm": 0.5327394008636475,
+      "learning_rate": 2.9315236947849213e-06,
+      "loss": 0.6432256698608398,
+      "step": 504
+    },
+    {
+      "epoch": 2.0159362549800797,
+      "grad_norm": 0.9132230281829834,
+      "learning_rate": 2.916733829610726e-06,
+      "loss": 0.6147190928459167,
+      "step": 506
+    },
+    {
+      "epoch": 2.0239043824701195,
+      "grad_norm": 0.10126911848783493,
+      "learning_rate": 2.901936716805302e-06,
+      "loss": 0.5483989715576172,
+      "step": 508
+    },
+    {
+      "epoch": 2.0318725099601593,
+      "grad_norm": 0.3067198097705841,
+      "learning_rate": 2.887132999573554e-06,
+      "loss": 0.8682755827903748,
+      "step": 510
+    },
+    {
+      "epoch": 2.039840637450199,
+      "grad_norm": 1.6026277542114258,
+      "learning_rate": 2.8723233214074675e-06,
+      "loss": 0.7569538950920105,
+      "step": 512
+    },
+    {
+      "epoch": 2.047808764940239,
+      "grad_norm": 1.0882881879806519,
+      "learning_rate": 2.857508326058142e-06,
+      "loss": 0.277357816696167,
+      "step": 514
+    },
+    {
+      "epoch": 2.055776892430279,
+      "grad_norm": 0.478473961353302,
+      "learning_rate": 2.842688657507805e-06,
+      "loss": 0.9134977459907532,
+      "step": 516
+    },
+    {
+      "epoch": 2.0637450199203187,
+      "grad_norm": 5.478097438812256,
+      "learning_rate": 2.8278649599418194e-06,
+      "loss": 0.5646745562553406,
+      "step": 518
+    },
+    {
+      "epoch": 2.0717131474103585,
+      "grad_norm": 0.9714582562446594,
+      "learning_rate": 2.813037877720683e-06,
+      "loss": 0.8388693928718567,
+      "step": 520
+    },
+    {
+      "epoch": 2.0796812749003983,
+      "grad_norm": 2.761141061782837,
+      "learning_rate": 2.798208055352017e-06,
+      "loss": 1.0104105472564697,
+      "step": 522
+    },
+    {
+      "epoch": 2.087649402390438,
+      "grad_norm": 0.9498922228813171,
+      "learning_rate": 2.7833761374625546e-06,
+      "loss": 0.9341850280761719,
+      "step": 524
+    },
+    {
+      "epoch": 2.095617529880478,
+      "grad_norm": 2.1572089195251465,
+      "learning_rate": 2.7685427687701172e-06,
+      "loss": 0.46415963768959045,
+      "step": 526
+    },
+    {
+      "epoch": 2.103585657370518,
+      "grad_norm": 2.22116756439209,
+      "learning_rate": 2.753708594055588e-06,
+      "loss": 0.777642548084259,
+      "step": 528
+    },
+    {
+      "epoch": 2.1115537848605577,
+      "grad_norm": 0.18673063814640045,
+      "learning_rate": 2.7388742581348893e-06,
+      "loss": 0.6769290566444397,
+      "step": 530
+    },
+    {
+      "epoch": 2.1195219123505975,
+      "grad_norm": 0.35420265793800354,
+      "learning_rate": 2.7240404058309487e-06,
+      "loss": 0.23003150522708893,
+      "step": 532
+    },
+    {
+      "epoch": 2.1274900398406373,
+      "grad_norm": 0.5746468305587769,
+      "learning_rate": 2.709207681945673e-06,
+      "loss": 0.7779282927513123,
+      "step": 534
+    },
+    {
+      "epoch": 2.135458167330677,
+      "grad_norm": 1.3902853727340698,
+      "learning_rate": 2.6943767312319167e-06,
+      "loss": 0.6417567133903503,
+      "step": 536
+    },
+    {
+      "epoch": 2.143426294820717,
+      "grad_norm": 0.4856451451778412,
+      "learning_rate": 2.679548198365462e-06,
+      "loss": 0.6673528552055359,
+      "step": 538
+    },
+    {
+      "epoch": 2.151394422310757,
+      "grad_norm": 0.6137409210205078,
+      "learning_rate": 2.6647227279169873e-06,
+      "loss": 0.9648311138153076,
+      "step": 540
+    },
+    {
+      "epoch": 2.1593625498007967,
+      "grad_norm": 0.8184354305267334,
+      "learning_rate": 2.6499009643240527e-06,
+      "loss": 0.9341200590133667,
+      "step": 542
+    },
+    {
+      "epoch": 2.1673306772908365,
+      "grad_norm": 0.4328656792640686,
+      "learning_rate": 2.6350835518630907e-06,
+      "loss": 0.7852800488471985,
+      "step": 544
+    },
+    {
+      "epoch": 2.1752988047808763,
+      "grad_norm": 0.7456241846084595,
+      "learning_rate": 2.620271134621397e-06,
+      "loss": 0.6618594527244568,
+      "step": 546
+    },
+    {
+      "epoch": 2.183266932270916,
+      "grad_norm": 0.8235044479370117,
+      "learning_rate": 2.605464356469131e-06,
+      "loss": 0.9851369261741638,
+      "step": 548
+    },
+    {
+      "epoch": 2.191235059760956,
+      "grad_norm": 0.37690049409866333,
+      "learning_rate": 2.5906638610313336e-06,
+      "loss": 1.0946252346038818,
+      "step": 550
+    },
+    {
+      "epoch": 2.199203187250996,
+      "grad_norm": 1.5211230516433716,
+      "learning_rate": 2.575870291659946e-06,
+      "loss": 0.7631247043609619,
+      "step": 552
+    },
+    {
+      "epoch": 2.2071713147410357,
+      "grad_norm": 1.3870434761047363,
+      "learning_rate": 2.5610842914058446e-06,
+      "loss": 0.4632687270641327,
+      "step": 554
+    },
+    {
+      "epoch": 2.2151394422310755,
+      "grad_norm": 0.4536760151386261,
+      "learning_rate": 2.5463065029908906e-06,
+      "loss": 0.6583999991416931,
+      "step": 556
+    },
+    {
+      "epoch": 2.2231075697211153,
+      "grad_norm": 1.5122045278549194,
+      "learning_rate": 2.5315375687799896e-06,
+      "loss": 0.8496344089508057,
+      "step": 558
+    },
+    {
+      "epoch": 2.231075697211155,
+      "grad_norm": 0.5606105923652649,
+      "learning_rate": 2.5167781307531714e-06,
+      "loss": 0.9522982835769653,
+      "step": 560
+    },
+    {
+      "epoch": 2.239043824701195,
+      "grad_norm": 0.5168128609657288,
+      "learning_rate": 2.5020288304776822e-06,
+      "loss": 1.0419187545776367,
+      "step": 562
+    },
+    {
+      "epoch": 2.247011952191235,
+      "grad_norm": 1.1527540683746338,
+      "learning_rate": 2.487290309080097e-06,
+      "loss": 0.8577864170074463,
+      "step": 564
+    },
+    {
+      "epoch": 2.2549800796812747,
+      "grad_norm": 0.5744875073432922,
+      "learning_rate": 2.472563207218454e-06,
+      "loss": 0.6748538017272949,
+      "step": 566
+    },
+    {
+      "epoch": 2.2629482071713145,
+      "grad_norm": 0.3853478729724884,
+      "learning_rate": 2.457848165054402e-06,
+      "loss": 0.7759252190589905,
+      "step": 568
+    },
+    {
+      "epoch": 2.2709163346613543,
+      "grad_norm": 0.6155991554260254,
+      "learning_rate": 2.4431458222253755e-06,
+      "loss": 1.0235545635223389,
+      "step": 570
+    },
+    {
+      "epoch": 2.278884462151394,
+      "grad_norm": 2.228321075439453,
+      "learning_rate": 2.4284568178167923e-06,
+      "loss": 0.6229550838470459,
+      "step": 572
+    },
+    {
+      "epoch": 2.2868525896414345,
+      "grad_norm": 0.41492176055908203,
+      "learning_rate": 2.413781790334272e-06,
+      "loss": 0.7239646315574646,
+      "step": 574
+    },
+    {
+      "epoch": 2.2948207171314743,
+      "grad_norm": 0.4200393557548523,
+      "learning_rate": 2.3991213776758776e-06,
+      "loss": 0.26977109909057617,
+      "step": 576
+    },
+    {
+      "epoch": 2.302788844621514,
+      "grad_norm": 0.847767174243927,
+      "learning_rate": 2.3844762171043975e-06,
+      "loss": 0.6452472805976868,
+      "step": 578
+    },
+    {
+      "epoch": 2.310756972111554,
+      "grad_norm": 0.5264169573783875,
+      "learning_rate": 2.3698469452196328e-06,
+      "loss": 0.6466808319091797,
+      "step": 580
+    },
+    {
+      "epoch": 2.318725099601594,
+      "grad_norm": 1.2575100660324097,
+      "learning_rate": 2.355234197930735e-06,
+      "loss": 0.6322447657585144,
+      "step": 582
+    },
+    {
+      "epoch": 2.3266932270916336,
+      "grad_norm": 1.5413649082183838,
+      "learning_rate": 2.340638610428554e-06,
+      "loss": 0.8036313056945801,
+      "step": 584
+    },
+    {
+      "epoch": 2.3346613545816735,
+      "grad_norm": 0.9299190640449524,
+      "learning_rate": 2.3260608171580412e-06,
+      "loss": 0.6360403895378113,
+      "step": 586
+    },
+    {
+      "epoch": 2.3426294820717133,
+      "grad_norm": 0.9121858477592468,
+      "learning_rate": 2.311501451790658e-06,
+      "loss": 0.8963300585746765,
+      "step": 588
+    },
+    {
+      "epoch": 2.350597609561753,
+      "grad_norm": 0.5690445899963379,
+      "learning_rate": 2.2969611471968386e-06,
+      "loss": 1.0182307958602905,
+      "step": 590
+    },
+    {
+      "epoch": 2.358565737051793,
+      "grad_norm": 0.5757805705070496,
+      "learning_rate": 2.2824405354184776e-06,
+      "loss": 1.0464677810668945,
+      "step": 592
+    },
+    {
+      "epoch": 2.366533864541833,
+      "grad_norm": 0.8780666589736938,
+      "learning_rate": 2.267940247641458e-06,
+      "loss": 0.7089881896972656,
+      "step": 594
+    },
+    {
+      "epoch": 2.3745019920318726,
+      "grad_norm": 0.35713064670562744,
+      "learning_rate": 2.2534609141682135e-06,
+      "loss": 1.0209615230560303,
+      "step": 596
+    },
+    {
+      "epoch": 2.3824701195219125,
+      "grad_norm": 0.2997743785381317,
+      "learning_rate": 2.2390031643903286e-06,
+      "loss": 0.7192252278327942,
+      "step": 598
+    },
+    {
+      "epoch": 2.3904382470119523,
+      "grad_norm": 1.2491768598556519,
+      "learning_rate": 2.224567626761185e-06,
+      "loss": 0.9660300612449646,
+      "step": 600
+    },
+    {
+      "epoch": 2.398406374501992,
+      "grad_norm": 0.5286229252815247,
+      "learning_rate": 2.2101549287686384e-06,
+      "loss": 1.030888557434082,
+      "step": 602
+    },
+    {
+      "epoch": 2.406374501992032,
+      "grad_norm": 3.877570629119873,
+      "learning_rate": 2.195765696907745e-06,
+      "loss": 0.5925008654594421,
+      "step": 604
+    },
+    {
+      "epoch": 2.414342629482072,
+      "grad_norm": 1.100961685180664,
+      "learning_rate": 2.1814005566535303e-06,
+      "loss": 0.23619602620601654,
+      "step": 606
+    },
+    {
+      "epoch": 2.4223107569721116,
+      "grad_norm": 0.25510329008102417,
+      "learning_rate": 2.167060132433798e-06,
+      "loss": 0.5397225618362427,
+      "step": 608
+    },
+    {
+      "epoch": 2.4302788844621515,
+      "grad_norm": 0.1959323287010193,
+      "learning_rate": 2.1527450476019895e-06,
+      "loss": 0.9447913765907288,
+      "step": 610
+    },
+    {
+      "epoch": 2.4382470119521913,
+      "grad_norm": 0.5705426335334778,
+      "learning_rate": 2.138455924410085e-06,
+      "loss": 0.6934602856636047,
+      "step": 612
+    },
+    {
+      "epoch": 2.446215139442231,
+      "grad_norm": 0.6084771752357483,
+      "learning_rate": 2.1241933839815606e-06,
+      "loss": 0.5833254456520081,
+      "step": 614
+    },
+    {
+      "epoch": 2.454183266932271,
+      "grad_norm": 0.446159690618515,
+      "learning_rate": 2.109958046284382e-06,
+      "loss": 0.9203718900680542,
+      "step": 616
+    },
+    {
+      "epoch": 2.462151394422311,
+      "grad_norm": 0.9361494779586792,
+      "learning_rate": 2.0957505301040604e-06,
+      "loss": 0.747937798500061,
+      "step": 618
+    },
+    {
+      "epoch": 2.4701195219123506,
+      "grad_norm": 1.2538261413574219,
+      "learning_rate": 2.081571453016756e-06,
+      "loss": 0.8260132670402527,
+      "step": 620
+    },
+    {
+      "epoch": 2.4780876494023905,
+      "grad_norm": 0.6002716422080994,
+      "learning_rate": 2.067421431362427e-06,
+      "loss": 0.8383951187133789,
+      "step": 622
+    },
+    {
+      "epoch": 2.4860557768924303,
+      "grad_norm": 1.006485939025879,
+      "learning_rate": 2.053301080218044e-06,
+      "loss": 0.7950708866119385,
+      "step": 624
+    },
+    {
+      "epoch": 2.49402390438247,
+      "grad_norm": 0.6845659613609314,
+      "learning_rate": 2.039211013370853e-06,
+      "loss": 0.6514892578125,
+      "step": 626
+    },
+    {
+      "epoch": 2.50199203187251,
+      "grad_norm": 0.5930602550506592,
+      "learning_rate": 2.025151843291693e-06,
+      "loss": 1.0060988664627075,
+      "step": 628
+    },
+    {
+      "epoch": 2.50996015936255,
+      "grad_norm": 0.6088513135910034,
+      "learning_rate": 2.0111241811083733e-06,
+      "loss": 0.4965975284576416,
+      "step": 630
+    },
+    {
+      "epoch": 2.5179282868525896,
+      "grad_norm": 1.4710122346878052,
+      "learning_rate": 1.9971286365791092e-06,
+      "loss": 0.4822719097137451,
+      "step": 632
+    },
+    {
+      "epoch": 2.5258964143426295,
+      "grad_norm": 0.5500730872154236,
+      "learning_rate": 1.9831658180660197e-06,
+      "loss": 1.0231213569641113,
+      "step": 634
+    },
+    {
+      "epoch": 2.5338645418326693,
+      "grad_norm": 1.0350205898284912,
+      "learning_rate": 1.9692363325086774e-06,
+      "loss": 0.9606418013572693,
+      "step": 636
+    },
+    {
+      "epoch": 2.541832669322709,
+      "grad_norm": 2.1699728965759277,
+      "learning_rate": 1.95534078539773e-06,
+      "loss": 1.017174243927002,
+      "step": 638
+    },
+    {
+      "epoch": 2.549800796812749,
+      "grad_norm": 0.5589404106140137,
+      "learning_rate": 1.9414797807485796e-06,
+      "loss": 0.7207890748977661,
+      "step": 640
+    },
+    {
+      "epoch": 2.557768924302789,
+      "grad_norm": 0.9904406070709229,
+      "learning_rate": 1.9276539210751295e-06,
+      "loss": 1.0014190673828125,
+      "step": 642
+    },
+    {
+      "epoch": 2.5657370517928286,
+      "grad_norm": 0.521212100982666,
+      "learning_rate": 1.9138638073635886e-06,
+      "loss": 1.094325304031372,
+      "step": 644
+    },
+    {
+      "epoch": 2.5737051792828685,
+      "grad_norm": 0.611664354801178,
+      "learning_rate": 1.90011003904635e-06,
+      "loss": 0.5898622870445251,
+      "step": 646
+    },
+    {
+      "epoch": 2.5816733067729083,
+      "grad_norm": 0.8709461688995361,
+      "learning_rate": 1.886393213975939e-06,
+      "loss": 1.0418039560317993,
+      "step": 648
+    },
+    {
+      "epoch": 2.589641434262948,
+      "grad_norm": 0.3867081105709076,
+      "learning_rate": 1.8727139283990203e-06,
+      "loss": 0.8018988966941833,
+      "step": 650
+    },
+    {
+      "epoch": 2.597609561752988,
+      "grad_norm": 4.007132053375244,
+      "learning_rate": 1.8590727769304789e-06,
+      "loss": 0.9010555148124695,
+      "step": 652
+    },
+    {
+      "epoch": 2.605577689243028,
+      "grad_norm": 0.6614624857902527,
+      "learning_rate": 1.8454703525275814e-06,
+      "loss": 0.6523377895355225,
+      "step": 654
+    },
+    {
+      "epoch": 2.6135458167330676,
+      "grad_norm": 0.836147129535675,
+      "learning_rate": 1.8319072464641912e-06,
+      "loss": 0.3493887782096863,
+      "step": 656
+    },
+    {
+      "epoch": 2.6215139442231075,
+      "grad_norm": 0.7081486582756042,
+      "learning_rate": 1.8183840483050741e-06,
+      "loss": 0.5639723539352417,
+      "step": 658
+    },
+    {
+      "epoch": 2.6294820717131473,
+      "grad_norm": 0.6342553496360779,
+      "learning_rate": 1.8049013458802667e-06,
+      "loss": 0.9494541883468628,
+      "step": 660
+    },
+    {
+      "epoch": 2.637450199203187,
+      "grad_norm": 1.1497447490692139,
+      "learning_rate": 1.7914597252595267e-06,
+      "loss": 0.9833539128303528,
+      "step": 662
+    },
+    {
+      "epoch": 2.645418326693227,
+      "grad_norm": 0.9240487217903137,
+      "learning_rate": 1.7780597707268582e-06,
+      "loss": 0.5901837944984436,
+      "step": 664
+    },
+    {
+      "epoch": 2.653386454183267,
+      "grad_norm": 0.785715639591217,
+      "learning_rate": 1.7647020647551105e-06,
+      "loss": 0.924850344657898,
+      "step": 666
+    },
+    {
+      "epoch": 2.6613545816733066,
+      "grad_norm": 0.1513298600912094,
+      "learning_rate": 1.7513871879806605e-06,
+      "loss": 0.5821530818939209,
+      "step": 668
+    },
+    {
+      "epoch": 2.6693227091633465,
+      "grad_norm": 1.1863737106323242,
+      "learning_rate": 1.7381157191781766e-06,
+      "loss": 0.7250882387161255,
+      "step": 670
+    },
+    {
+      "epoch": 2.6772908366533863,
+      "grad_norm": 0.3311910927295685,
+      "learning_rate": 1.7248882352354571e-06,
+      "loss": 0.3806076943874359,
+      "step": 672
+    },
+    {
+      "epoch": 2.685258964143426,
+      "grad_norm": 0.4173329472541809,
+      "learning_rate": 1.7117053111283528e-06,
+      "loss": 1.0334678888320923,
+      "step": 674
+    },
+    {
+      "epoch": 2.6932270916334664,
+      "grad_norm": 4.259430885314941,
+      "learning_rate": 1.6985675198957752e-06,
+      "loss": 0.6853829622268677,
+      "step": 676
+    },
+    {
+      "epoch": 2.7011952191235062,
+      "grad_norm": 0.539608359336853,
+      "learning_rate": 1.6854754326147916e-06,
+      "loss": 0.6058146357536316,
+      "step": 678
+    },
+    {
+      "epoch": 2.709163346613546,
+      "grad_norm": 0.42629122734069824,
+      "learning_rate": 1.6724296183757925e-06,
+      "loss": 1.0257009267807007,
+      "step": 680
+    },
+    {
+      "epoch": 2.717131474103586,
+      "grad_norm": 0.8992399573326111,
+      "learning_rate": 1.6594306442577629e-06,
+      "loss": 1.0538326501846313,
+      "step": 682
+    },
+    {
+      "epoch": 2.7250996015936257,
+      "grad_norm": 0.5011391639709473,
+      "learning_rate": 1.6464790753036298e-06,
+      "loss": 0.6967548727989197,
+      "step": 684
+    },
+    {
+      "epoch": 2.7330677290836656,
+      "grad_norm": 1.0441275835037231,
+      "learning_rate": 1.6335754744956935e-06,
+      "loss": 0.4877105951309204,
+      "step": 686
+    },
+    {
+      "epoch": 2.7410358565737054,
+      "grad_norm": 0.7679535746574402,
+      "learning_rate": 1.6207204027311685e-06,
+      "loss": 1.0346051454544067,
+      "step": 688
+    },
+    {
+      "epoch": 2.7490039840637452,
+      "grad_norm": 0.9672122001647949,
+      "learning_rate": 1.6079144187977925e-06,
+      "loss": 0.5439547896385193,
+      "step": 690
+    },
+    {
+      "epoch": 2.756972111553785,
+      "grad_norm": 0.5946493148803711,
+      "learning_rate": 1.595158079349542e-06,
+      "loss": 1.1041629314422607,
+      "step": 692
+    },
+    {
+      "epoch": 2.764940239043825,
+      "grad_norm": 3.5004076957702637,
+      "learning_rate": 1.5824519388824311e-06,
+      "loss": 0.8648486137390137,
+      "step": 694
+    },
+    {
+      "epoch": 2.7729083665338647,
+      "grad_norm": 1.398216962814331,
+      "learning_rate": 1.5697965497104117e-06,
+      "loss": 0.669555127620697,
+      "step": 696
+    },
+    {
+      "epoch": 2.7808764940239046,
+      "grad_norm": 0.5257605910301208,
+      "learning_rate": 1.557192461941367e-06,
+      "loss": 0.9388008713722229,
+      "step": 698
+    },
+    {
+      "epoch": 2.7888446215139444,
+      "grad_norm": 0.6866509914398193,
+      "learning_rate": 1.5446402234531934e-06,
+      "loss": 1.0469776391983032,
+      "step": 700
+    },
+    {
+      "epoch": 2.7968127490039842,
+      "grad_norm": 0.9670447111129761,
+      "learning_rate": 1.5321403798699924e-06,
+      "loss": 0.6402778625488281,
+      "step": 702
+    },
+    {
+      "epoch": 2.804780876494024,
+      "grad_norm": 0.7777611613273621,
+      "learning_rate": 1.5196934745383456e-06,
+      "loss": 1.2216871976852417,
+      "step": 704
+    },
+    {
+      "epoch": 2.812749003984064,
+      "grad_norm": 0.4490237236022949,
+      "learning_rate": 1.5073000485037044e-06,
+      "loss": 1.2219018936157227,
+      "step": 706
+    },
+    {
+      "epoch": 2.8207171314741037,
+      "grad_norm": 0.550182580947876,
+      "learning_rate": 1.4949606404868638e-06,
+      "loss": 1.094996690750122,
+      "step": 708
+    },
+    {
+      "epoch": 2.8286852589641436,
+      "grad_norm": 0.6918236613273621,
+      "learning_rate": 1.4826757868605519e-06,
+      "loss": 0.5505000948905945,
+      "step": 710
+    },
+    {
+      "epoch": 2.8366533864541834,
+      "grad_norm": 0.626388669013977,
+      "learning_rate": 1.4704460216261125e-06,
+      "loss": 1.0855085849761963,
+      "step": 712
+    },
+    {
+      "epoch": 2.8446215139442232,
+      "grad_norm": 0.977725088596344,
+      "learning_rate": 1.458271876390287e-06,
+      "loss": 0.6149780750274658,
+      "step": 714
+    },
+    {
+      "epoch": 2.852589641434263,
+      "grad_norm": 0.8420648574829102,
+      "learning_rate": 1.446153880342117e-06,
+      "loss": 1.0308772325515747,
+      "step": 716
+    },
+    {
+      "epoch": 2.860557768924303,
+      "grad_norm": 2.214899778366089,
+      "learning_rate": 1.4340925602299337e-06,
+      "loss": 0.636957049369812,
+      "step": 718
+    },
+    {
+      "epoch": 2.8685258964143427,
+      "grad_norm": 0.5352880358695984,
+      "learning_rate": 1.4220884403384656e-06,
+      "loss": 0.5828282833099365,
+      "step": 720
+    },
+    {
+      "epoch": 2.8764940239043826,
+      "grad_norm": 0.5229229927062988,
+      "learning_rate": 1.4101420424660433e-06,
+      "loss": 0.7304311990737915,
+      "step": 722
+    },
+    {
+      "epoch": 2.8844621513944224,
+      "grad_norm": 0.3471089005470276,
+      "learning_rate": 1.3982538859019219e-06,
+      "loss": 0.614575982093811,
+      "step": 724
+    },
+    {
+      "epoch": 2.8924302788844622,
+      "grad_norm": 0.33785372972488403,
+      "learning_rate": 1.3864244874037096e-06,
+      "loss": 0.9909072518348694,
+      "step": 726
+    },
+    {
+      "epoch": 2.900398406374502,
+      "grad_norm": 0.5180720090866089,
+      "learning_rate": 1.3746543611749007e-06,
+      "loss": 0.5618436336517334,
+      "step": 728
+    },
+    {
+      "epoch": 2.908366533864542,
+      "grad_norm": 0.43437862396240234,
+      "learning_rate": 1.3629440188425285e-06,
+      "loss": 0.9497717618942261,
+      "step": 730
+    },
+    {
+      "epoch": 2.9163346613545817,
+      "grad_norm": 3.7571897506713867,
+      "learning_rate": 1.351293969434921e-06,
+      "loss": 0.3881228268146515,
+      "step": 732
+    },
+    {
+      "epoch": 2.9243027888446216,
+      "grad_norm": 0.49480491876602173,
+      "learning_rate": 1.3397047193595809e-06,
+      "loss": 1.0534650087356567,
+      "step": 734
+    },
+    {
+      "epoch": 2.9322709163346614,
+      "grad_norm": 2.111862897872925,
+      "learning_rate": 1.3281767723811651e-06,
+      "loss": 0.4529063403606415,
+      "step": 736
+    },
+    {
+      "epoch": 2.9402390438247012,
+      "grad_norm": 1.6681751012802124,
+      "learning_rate": 1.3167106295995923e-06,
+      "loss": 0.8840406537055969,
+      "step": 738
+    },
+    {
+      "epoch": 2.948207171314741,
+      "grad_norm": 0.38326793909072876,
+      "learning_rate": 1.3053067894282624e-06,
+      "loss": 1.157111644744873,
+      "step": 740
+    },
+    {
+      "epoch": 2.956175298804781,
+      "grad_norm": 0.9434833526611328,
+      "learning_rate": 1.2939657475723844e-06,
+      "loss": 0.3580433130264282,
+      "step": 742
+    },
+    {
+      "epoch": 2.9641434262948207,
+      "grad_norm": 0.3832603096961975,
+      "learning_rate": 1.2826879970074346e-06,
+      "loss": 0.9365513920783997,
+      "step": 744
+    },
+    {
+      "epoch": 2.9721115537848606,
+      "grad_norm": 0.39480799436569214,
+      "learning_rate": 1.271474027957727e-06,
+      "loss": 0.8868306279182434,
+      "step": 746
+    },
+    {
+      "epoch": 2.9800796812749004,
+      "grad_norm": 0.4570213258266449,
+      "learning_rate": 1.2603243278751043e-06,
+      "loss": 0.5333258509635925,
+      "step": 748
+    },
+    {
+      "epoch": 2.9880478087649402,
+      "grad_norm": 0.8246273994445801,
+      "learning_rate": 1.249239381417746e-06,
+      "loss": 0.12774328887462616,
+      "step": 750
+    },
+    {
+      "epoch": 2.99601593625498,
+      "grad_norm": 0.6178113222122192,
+      "learning_rate": 1.238219670429104e-06,
+      "loss": 1.019992470741272,
+      "step": 752
+    },
+    {
+      "epoch": 3.00398406374502,
+      "grad_norm": 0.5595139861106873,
+      "learning_rate": 1.2272656739169589e-06,
+      "loss": 0.579146146774292,
+      "step": 754
+    },
+    {
+      "epoch": 3.0119521912350598,
+      "grad_norm": 0.5749949812889099,
+      "learning_rate": 1.2163778680325948e-06,
+      "loss": 0.625841498374939,
+      "step": 756
+    },
+    {
+      "epoch": 3.0199203187250996,
+      "grad_norm": 0.6717772483825684,
+      "learning_rate": 1.2055567260501055e-06,
+      "loss": 0.3705125153064728,
+      "step": 758
+    },
+    {
+      "epoch": 3.0278884462151394,
+      "grad_norm": 0.5492572784423828,
+      "learning_rate": 1.1948027183458182e-06,
+      "loss": 0.6853355765342712,
+      "step": 760
+    },
+    {
+      "epoch": 3.0358565737051793,
+      "grad_norm": 0.6478875279426575,
+      "learning_rate": 1.18411631237785e-06,
+      "loss": 0.521458625793457,
+      "step": 762
+    },
+    {
+      "epoch": 3.043824701195219,
+      "grad_norm": 0.36975279450416565,
+      "learning_rate": 1.173497972665788e-06,
+      "loss": 0.9694585204124451,
+      "step": 764
+    },
+    {
+      "epoch": 3.051792828685259,
+      "grad_norm": 0.2931518256664276,
+      "learning_rate": 1.1629481607704964e-06,
+      "loss": 0.1518229842185974,
+      "step": 766
+    },
+    {
+      "epoch": 3.0597609561752988,
+      "grad_norm": 0.37477609515190125,
+      "learning_rate": 1.1524673352740543e-06,
+      "loss": 0.5745505094528198,
+      "step": 768
+    },
+    {
+      "epoch": 3.0677290836653386,
+      "grad_norm": 0.5869929194450378,
+      "learning_rate": 1.1420559517598206e-06,
+      "loss": 0.8037846684455872,
+      "step": 770
+    },
+    {
+      "epoch": 3.0756972111553784,
+      "grad_norm": 0.4266442060470581,
+      "learning_rate": 1.13171446279263e-06,
+      "loss": 0.9616826772689819,
+      "step": 772
+    },
+    {
+      "epoch": 3.0836653386454183,
+      "grad_norm": 1.799504041671753,
+      "learning_rate": 1.1214433178991258e-06,
+      "loss": 0.8843712210655212,
+      "step": 774
+    },
+    {
+      "epoch": 3.091633466135458,
+      "grad_norm": 0.18844708800315857,
+      "learning_rate": 1.111242963548212e-06,
+      "loss": 0.558309018611908,
+      "step": 776
+    },
+    {
+      "epoch": 3.099601593625498,
+      "grad_norm": 0.6273339986801147,
+      "learning_rate": 1.1011138431316538e-06,
+      "loss": 0.5815000534057617,
+      "step": 778
+    },
+    {
+      "epoch": 3.1075697211155378,
+      "grad_norm": 0.540294885635376,
+      "learning_rate": 1.0910563969447969e-06,
+      "loss": 0.9752479195594788,
+      "step": 780
+    },
+    {
+      "epoch": 3.1155378486055776,
+      "grad_norm": 0.23366230726242065,
+      "learning_rate": 1.0810710621674355e-06,
+      "loss": 0.2870320975780487,
+      "step": 782
+    },
+    {
+      "epoch": 3.1235059760956174,
+      "grad_norm": 0.4389194846153259,
+      "learning_rate": 1.0711582728448034e-06,
+      "loss": 1.0099979639053345,
+      "step": 784
+    },
+    {
+      "epoch": 3.1314741035856573,
+      "grad_norm": 0.5615717768669128,
+      "learning_rate": 1.0613184598687088e-06,
+      "loss": 0.791538417339325,
+      "step": 786
+    },
+    {
+      "epoch": 3.139442231075697,
+      "grad_norm": 0.9747589826583862,
+      "learning_rate": 1.0515520509588076e-06,
+      "loss": 0.5976436734199524,
+      "step": 788
+    },
+    {
+      "epoch": 3.147410358565737,
+      "grad_norm": 0.24722778797149658,
+      "learning_rate": 1.0418594706440022e-06,
+      "loss": 0.5254042744636536,
+      "step": 790
+    },
+    {
+      "epoch": 3.1553784860557768,
+      "grad_norm": 1.2574716806411743,
+      "learning_rate": 1.0322411402439982e-06,
+      "loss": 0.7492045164108276,
+      "step": 792
+    },
+    {
+      "epoch": 3.1633466135458166,
+      "grad_norm": 0.08566952496767044,
+      "learning_rate": 1.0226974778509846e-06,
+      "loss": 0.23793993890285492,
+      "step": 794
+    },
+    {
+      "epoch": 3.1713147410358564,
+      "grad_norm": 0.915134847164154,
+      "learning_rate": 1.0132288983114611e-06,
+      "loss": 0.6315504312515259,
+      "step": 796
+    },
+    {
+      "epoch": 3.1792828685258963,
+      "grad_norm": 1.230481505393982,
+      "learning_rate": 1.003835813208206e-06,
+      "loss": 0.5491132140159607,
+      "step": 798
+    },
+    {
+      "epoch": 3.187250996015936,
+      "grad_norm": 0.957248330116272,
+      "learning_rate": 9.945186308423832e-07,
+      "loss": 0.8668388724327087,
+      "step": 800
+    },
+    {
+      "epoch": 3.195219123505976,
+      "grad_norm": 0.26860058307647705,
+      "learning_rate": 9.852777562157983e-07,
+      "loss": 0.7234906554222107,
+      "step": 802
+    },
+    {
+      "epoch": 3.2031872509960158,
+      "grad_norm": 1.195494294166565,
+      "learning_rate": 9.7611359101329e-07,
+      "loss": 0.6418740153312683,
+      "step": 804
+    },
+    {
+      "epoch": 3.2111553784860556,
+      "grad_norm": 1.2842307090759277,
+      "learning_rate": 9.670265335852725e-07,
+      "loss": 1.1286014318466187,
+      "step": 806
+    },
+    {
+      "epoch": 3.2191235059760954,
+      "grad_norm": 1.821799397468567,
+      "learning_rate": 9.580169789304152e-07,
+      "loss": 0.39762043952941895,
+      "step": 808
+    },
+    {
+      "epoch": 3.2270916334661353,
+      "grad_norm": 0.583942711353302,
+      "learning_rate": 9.490853186784801e-07,
+      "loss": 0.7078452110290527,
+      "step": 810
+    },
+    {
+      "epoch": 3.235059760956175,
+      "grad_norm": 1.0375887155532837,
+      "learning_rate": 9.402319410732911e-07,
+      "loss": 0.37614575028419495,
+      "step": 812
+    },
+    {
+      "epoch": 3.243027888446215,
+      "grad_norm": 0.3879964053630829,
+      "learning_rate": 9.314572309558617e-07,
+      "loss": 0.8332592248916626,
+      "step": 814
+    },
+    {
+      "epoch": 3.2509960159362548,
+      "grad_norm": 14.581340789794922,
+      "learning_rate": 9.227615697476674e-07,
+      "loss": 0.41820454597473145,
+      "step": 816
+    },
+    {
+      "epoch": 3.2589641434262946,
+      "grad_norm": 0.8184705376625061,
+      "learning_rate": 9.141453354340619e-07,
+      "loss": 0.9523636102676392,
+      "step": 818
+    },
+    {
+      "epoch": 3.2669322709163344,
+      "grad_norm": 2.0967276096343994,
+      "learning_rate": 9.056089025478495e-07,
+      "loss": 0.4982432723045349,
+      "step": 820
+    },
+    {
+      "epoch": 3.2749003984063743,
+      "grad_norm": 1.5760488510131836,
+      "learning_rate": 8.971526421530067e-07,
+      "loss": 0.7282808423042297,
+      "step": 822
+    },
+    {
+      "epoch": 3.2828685258964145,
+      "grad_norm": 0.9606645107269287,
+      "learning_rate": 8.887769218285469e-07,
+      "loss": 0.5421037673950195,
+      "step": 824
+    },
+    {
+      "epoch": 3.2908366533864544,
+      "grad_norm": 0.4770655035972595,
+      "learning_rate": 8.804821056525498e-07,
+      "loss": 0.9671206474304199,
+      "step": 826
+    },
+    {
+      "epoch": 3.298804780876494,
+      "grad_norm": 0.4865878224372864,
+      "learning_rate": 8.722685541863284e-07,
+      "loss": 0.9929076433181763,
+      "step": 828
+    },
+    {
+      "epoch": 3.306772908366534,
+      "grad_norm": 0.7179232835769653,
+      "learning_rate": 8.641366244587612e-07,
+      "loss": 0.8825250864028931,
+      "step": 830
+    },
+    {
+      "epoch": 3.314741035856574,
+      "grad_norm": 1.7425504922866821,
+      "learning_rate": 8.560866699507703e-07,
+      "loss": 0.5349132418632507,
+      "step": 832
+    },
+    {
+      "epoch": 3.3227091633466137,
+      "grad_norm": 1.3007465600967407,
+      "learning_rate": 8.481190405799569e-07,
+      "loss": 0.3397764563560486,
+      "step": 834
+    },
+    {
+      "epoch": 3.3306772908366535,
+      "grad_norm": 2.9653308391571045,
+      "learning_rate": 8.402340826853908e-07,
+      "loss": 0.6012100577354431,
+      "step": 836
+    },
+    {
+      "epoch": 3.3386454183266934,
+      "grad_norm": 3.8056063652038574,
+      "learning_rate": 8.324321390125548e-07,
+      "loss": 0.8524638414382935,
+      "step": 838
+    },
+    {
+      "epoch": 3.346613545816733,
+      "grad_norm": 1.1574907302856445,
+      "learning_rate": 8.247135486984487e-07,
+      "loss": 0.19105130434036255,
+      "step": 840
+    },
+    {
+      "epoch": 3.354581673306773,
+      "grad_norm": 1.3461089134216309,
+      "learning_rate": 8.170786472568451e-07,
+      "loss": 0.7022271156311035,
+      "step": 842
+    },
+    {
+      "epoch": 3.362549800796813,
+      "grad_norm": 0.4820784032344818,
+      "learning_rate": 8.095277665637061e-07,
+      "loss": 0.6833860874176025,
+      "step": 844
+    },
+    {
+      "epoch": 3.3705179282868527,
+      "grad_norm": 0.6242122650146484,
+      "learning_rate": 8.020612348427572e-07,
+      "loss": 0.9958716630935669,
+      "step": 846
+    },
+    {
+      "epoch": 3.3784860557768925,
+      "grad_norm": 0.5448309779167175,
+      "learning_rate": 7.946793766512195e-07,
+      "loss": 0.9622980356216431,
+      "step": 848
+    },
+    {
+      "epoch": 3.3864541832669324,
+      "grad_norm": 1.0548478364944458,
+      "learning_rate": 7.873825128657034e-07,
+      "loss": 0.9945654273033142,
+      "step": 850
+    },
+    {
+      "epoch": 3.394422310756972,
+      "grad_norm": 0.40971457958221436,
+      "learning_rate": 7.801709606682573e-07,
+      "loss": 0.9394071102142334,
+      "step": 852
+    },
+    {
+      "epoch": 3.402390438247012,
+      "grad_norm": 0.770642101764679,
+      "learning_rate": 7.730450335325852e-07,
+      "loss": 0.5802016854286194,
+      "step": 854
+    },
+    {
+      "epoch": 3.410358565737052,
+      "grad_norm": 0.8857874870300293,
+      "learning_rate": 7.660050412104153e-07,
+      "loss": 0.8953830003738403,
+      "step": 856
+    },
+    {
+      "epoch": 3.4183266932270917,
+      "grad_norm": 0.7635868191719055,
+      "learning_rate": 7.5905128971804e-07,
+      "loss": 1.06135892868042,
+      "step": 858
+    },
+    {
+      "epoch": 3.4262948207171315,
+      "grad_norm": 0.5823217034339905,
+      "learning_rate": 7.521840813230098e-07,
+      "loss": 1.0106362104415894,
+      "step": 860
+    },
+    {
+      "epoch": 3.4342629482071714,
+      "grad_norm": 0.4080336093902588,
+      "learning_rate": 7.454037145309986e-07,
+      "loss": 0.692585289478302,
+      "step": 862
+    },
+    {
+      "epoch": 3.442231075697211,
+      "grad_norm": 0.9530379176139832,
+      "learning_rate": 7.38710484072826e-07,
+      "loss": 0.536604642868042,
+      "step": 864
+    },
+    {
+      "epoch": 3.450199203187251,
+      "grad_norm": 0.512589156627655,
+      "learning_rate": 7.321046808916427e-07,
+      "loss": 0.8482944369316101,
+      "step": 866
+    },
+    {
+      "epoch": 3.458167330677291,
+      "grad_norm": 1.1212929487228394,
+      "learning_rate": 7.255865921302902e-07,
+      "loss": 0.7135361433029175,
+      "step": 868
+    },
+    {
+      "epoch": 3.4661354581673307,
+      "grad_norm": 0.48777833580970764,
+      "learning_rate": 7.191565011188149e-07,
+      "loss": 0.5262725353240967,
+      "step": 870
+    },
+    {
+      "epoch": 3.4741035856573705,
+      "grad_norm": 0.12542793154716492,
+      "learning_rate": 7.128146873621533e-07,
+      "loss": 0.5754909515380859,
+      "step": 872
+    },
+    {
+      "epoch": 3.4820717131474104,
+      "grad_norm": 0.2517932951450348,
+      "learning_rate": 7.065614265279819e-07,
+      "loss": 0.694409191608429,
+      "step": 874
+    },
+    {
+      "epoch": 3.49003984063745,
+      "grad_norm": 0.600282609462738,
+      "learning_rate": 7.00396990434734e-07,
+      "loss": 0.5396724343299866,
+      "step": 876
+    },
+    {
+      "epoch": 3.49800796812749,
+      "grad_norm": 0.7411002516746521,
+      "learning_rate": 6.943216470397872e-07,
+      "loss": 0.5474315881729126,
+      "step": 878
+    },
+    {
+      "epoch": 3.50597609561753,
+      "grad_norm": 0.1864290088415146,
+      "learning_rate": 6.883356604278116e-07,
+      "loss": 0.41033685207366943,
+      "step": 880
+    },
+    {
+      "epoch": 3.5139442231075697,
+      "grad_norm": 0.47263798117637634,
+      "learning_rate": 6.824392907992938e-07,
+      "loss": 0.6487886309623718,
+      "step": 882
+    },
+    {
+      "epoch": 3.5219123505976095,
+      "grad_norm": 1.165248155593872,
+      "learning_rate": 6.76632794459224e-07,
+      "loss": 0.17825782299041748,
+      "step": 884
+    },
+    {
+      "epoch": 3.5298804780876494,
+      "grad_norm": 1.130654215812683,
+      "learning_rate": 6.709164238059568e-07,
+      "loss": 0.7878249287605286,
+      "step": 886
+    },
+    {
+      "epoch": 3.537848605577689,
+      "grad_norm": 0.940370500087738,
+      "learning_rate": 6.652904273202385e-07,
+      "loss": 0.9679273962974548,
+      "step": 888
+    },
+    {
+      "epoch": 3.545816733067729,
+      "grad_norm": 0.4817887544631958,
+      "learning_rate": 6.597550495544071e-07,
+      "loss": 0.49837806820869446,
+      "step": 890
+    },
+    {
+      "epoch": 3.553784860557769,
+      "grad_norm": 1.3066110610961914,
+      "learning_rate": 6.54310531121762e-07,
+      "loss": 0.9139914512634277,
+      "step": 892
+    },
+    {
+      "epoch": 3.5617529880478087,
+      "grad_norm": 0.49228933453559875,
+      "learning_rate": 6.489571086861027e-07,
+      "loss": 1.0401159524917603,
+      "step": 894
+    },
+    {
+      "epoch": 3.5697211155378485,
+      "grad_norm": 0.27108216285705566,
+      "learning_rate": 6.436950149514448e-07,
+      "loss": 0.22903943061828613,
+      "step": 896
+    },
+    {
+      "epoch": 3.5776892430278884,
+      "grad_norm": 0.3529016077518463,
+      "learning_rate": 6.385244786519036e-07,
+      "loss": 0.6440197229385376,
+      "step": 898
+    },
+    {
+      "epoch": 3.585657370517928,
+      "grad_norm": 0.6561027765274048,
+      "learning_rate": 6.334457245417491e-07,
+      "loss": 0.23832546174526215,
+      "step": 900
+    },
+    {
+      "epoch": 3.593625498007968,
+      "grad_norm": 2.001147508621216,
+      "learning_rate": 6.284589733856403e-07,
+      "loss": 0.10949621349573135,
+      "step": 902
+    },
+    {
+      "epoch": 3.601593625498008,
+      "grad_norm": 1.1177871227264404,
+      "learning_rate": 6.235644419490253e-07,
+      "loss": 0.7543700337409973,
+      "step": 904
+    },
+    {
+      "epoch": 3.6095617529880477,
+      "grad_norm": 0.6228533387184143,
+      "learning_rate": 6.18762342988722e-07,
+      "loss": 0.6378473043441772,
+      "step": 906
+    },
+    {
+      "epoch": 3.6175298804780875,
+      "grad_norm": 0.7163503766059875,
+      "learning_rate": 6.140528852436666e-07,
+      "loss": 0.8991875052452087,
+      "step": 908
+    },
+    {
+      "epoch": 3.6254980079681274,
+      "grad_norm": 1.8865348100662231,
+      "learning_rate": 6.094362734258438e-07,
+      "loss": 0.876941442489624,
+      "step": 910
+    },
+    {
+      "epoch": 3.633466135458167,
+      "grad_norm": 0.514621376991272,
+      "learning_rate": 6.049127082113853e-07,
+      "loss": 0.04432225227355957,
+      "step": 912
+    },
+    {
+      "epoch": 3.641434262948207,
+      "grad_norm": 0.35589325428009033,
+      "learning_rate": 6.004823862318475e-07,
+      "loss": 0.8069411516189575,
+      "step": 914
+    },
+    {
+      "epoch": 3.649402390438247,
+      "grad_norm": 1.377711296081543,
+      "learning_rate": 5.961455000656656e-07,
+      "loss": 0.8312397599220276,
+      "step": 916
+    },
+    {
+      "epoch": 3.6573705179282867,
+      "grad_norm": 2.0273678302764893,
+      "learning_rate": 5.919022382297808e-07,
+      "loss": 0.19391275942325592,
+      "step": 918
+    },
+    {
+      "epoch": 3.6653386454183265,
+      "grad_norm": 0.7216324806213379,
+      "learning_rate": 5.877527851714479e-07,
+      "loss": 0.9630604386329651,
+      "step": 920
+    },
+    {
+      "epoch": 3.6733067729083664,
+      "grad_norm": 1.1258862018585205,
+      "learning_rate": 5.836973212602144e-07,
+      "loss": 0.14493830502033234,
+      "step": 922
+    },
+    {
+      "epoch": 3.681274900398406,
+      "grad_norm": 2.309058427810669,
+      "learning_rate": 5.797360227800832e-07,
+      "loss": 0.9860643744468689,
+      "step": 924
+    },
+    {
+      "epoch": 3.6892430278884465,
+      "grad_norm": 0.4696836471557617,
+      "learning_rate": 5.758690619218493e-07,
+      "loss": 0.2675930857658386,
+      "step": 926
+    },
+    {
+      "epoch": 3.6972111553784863,
+      "grad_norm": 0.6530823707580566,
+      "learning_rate": 5.720966067756131e-07,
+      "loss": 0.5636778473854065,
+      "step": 928
+    },
+    {
+      "epoch": 3.705179282868526,
+      "grad_norm": 0.2877671718597412,
+      "learning_rate": 5.684188213234763e-07,
+      "loss": 0.8574612140655518,
+      "step": 930
+    },
+    {
+      "epoch": 3.713147410358566,
+      "grad_norm": 0.49212750792503357,
+      "learning_rate": 5.648358654324119e-07,
+      "loss": 0.9533630609512329,
+      "step": 932
+    },
+    {
+      "epoch": 3.721115537848606,
+      "grad_norm": 1.5291314125061035,
+      "learning_rate": 5.613478948473166e-07,
+      "loss": 0.4991132318973541,
+      "step": 934
+    },
+    {
+      "epoch": 3.7290836653386457,
+      "grad_norm": 0.3981715142726898,
+      "learning_rate": 5.579550611842395e-07,
+      "loss": 0.87935870885849,
+      "step": 936
+    },
+    {
+      "epoch": 3.7370517928286855,
+      "grad_norm": 1.3286118507385254,
+      "learning_rate": 5.546575119237923e-07,
+      "loss": 0.656298816204071,
+      "step": 938
+    },
+    {
+      "epoch": 3.7450199203187253,
+      "grad_norm": 0.6346187591552734,
+      "learning_rate": 5.514553904047394e-07,
+      "loss": 0.8706903457641602,
+      "step": 940
+    },
+    {
+      "epoch": 3.752988047808765,
+      "grad_norm": 1.8346360921859741,
+      "learning_rate": 5.483488358177647e-07,
+      "loss": 0.731263279914856,
+      "step": 942
+    },
+    {
+      "epoch": 3.760956175298805,
+      "grad_norm": 0.47826090455055237,
+      "learning_rate": 5.453379831994238e-07,
+      "loss": 0.5193066000938416,
+      "step": 944
+    },
+    {
+      "epoch": 3.768924302788845,
+      "grad_norm": 0.3809550702571869,
+      "learning_rate": 5.424229634262735e-07,
+      "loss": 0.9831307530403137,
+      "step": 946
+    },
+    {
+      "epoch": 3.7768924302788847,
+      "grad_norm": 0.5871403813362122,
+      "learning_rate": 5.396039032091826e-07,
+      "loss": 0.5808805823326111,
+      "step": 948
+    },
+    {
+      "epoch": 3.7848605577689245,
+      "grad_norm": 0.5625377893447876,
+      "learning_rate": 5.368809250878228e-07,
+      "loss": 0.9560198187828064,
+      "step": 950
+    },
+    {
+      "epoch": 3.7928286852589643,
+      "grad_norm": 0.4426833987236023,
+      "learning_rate": 5.342541474253438e-07,
+      "loss": 0.9622094035148621,
+      "step": 952
+    },
+    {
+      "epoch": 3.800796812749004,
+      "grad_norm": 1.08371901512146,
+      "learning_rate": 5.317236844032285e-07,
+      "loss": 0.5493331551551819,
+      "step": 954
+    },
+    {
+      "epoch": 3.808764940239044,
+      "grad_norm": 0.965507447719574,
+      "learning_rate": 5.292896460163282e-07,
+      "loss": 0.768927812576294,
+      "step": 956
+    },
+    {
+      "epoch": 3.816733067729084,
+      "grad_norm": 2.2096176147460938,
+      "learning_rate": 5.269521380680823e-07,
+      "loss": 0.5867412686347961,
+      "step": 958
+    },
+    {
+      "epoch": 3.8247011952191237,
+      "grad_norm": 1.0418481826782227,
+      "learning_rate": 5.247112621659186e-07,
+      "loss": 0.7000386714935303,
+      "step": 960
+    },
+    {
+      "epoch": 3.8326693227091635,
+      "grad_norm": 0.31872454285621643,
+      "learning_rate": 5.225671157168378e-07,
+      "loss": 0.559232234954834,
+      "step": 962
+    },
+    {
+      "epoch": 3.8406374501992033,
+      "grad_norm": 0.6209536194801331,
+      "learning_rate": 5.205197919231773e-07,
+      "loss": 0.9737389087677002,
+      "step": 964
+    },
+    {
+      "epoch": 3.848605577689243,
+      "grad_norm": 6.848320960998535,
+      "learning_rate": 5.185693797785624e-07,
+      "loss": 0.37415996193885803,
+      "step": 966
+    },
+    {
+      "epoch": 3.856573705179283,
+      "grad_norm": 0.5784829258918762,
+      "learning_rate": 5.167159640640358e-07,
+      "loss": 0.5589777231216431,
+      "step": 968
+    },
+    {
+      "epoch": 3.864541832669323,
+      "grad_norm": 1.4430149793624878,
+      "learning_rate": 5.149596253443733e-07,
+      "loss": 0.8073258996009827,
+      "step": 970
+    },
+    {
+      "epoch": 3.8725099601593627,
+      "grad_norm": 0.1828831285238266,
+      "learning_rate": 5.133004399645818e-07,
+      "loss": 0.548916220664978,
+      "step": 972
+    },
+    {
+      "epoch": 3.8804780876494025,
+      "grad_norm": 0.6177156567573547,
+      "learning_rate": 5.117384800465806e-07,
+      "loss": 0.45797914266586304,
+      "step": 974
+    },
+    {
+      "epoch": 3.8884462151394423,
+      "grad_norm": 0.43884581327438354,
+      "learning_rate": 5.102738134860656e-07,
+      "loss": 0.9594522714614868,
+      "step": 976
+    },
+    {
+      "epoch": 3.896414342629482,
+      "grad_norm": 3.0693905353546143,
+      "learning_rate": 5.089065039495598e-07,
+      "loss": 0.5316993594169617,
+      "step": 978
+    },
+    {
+      "epoch": 3.904382470119522,
+      "grad_norm": 0.33187830448150635,
+      "learning_rate": 5.076366108716438e-07,
+      "loss": 0.3797495365142822,
+      "step": 980
+    },
+    {
+      "epoch": 3.912350597609562,
+      "grad_norm": 0.8389091491699219,
+      "learning_rate": 5.064641894523737e-07,
+      "loss": 0.47458699345588684,
+      "step": 982
+    },
+    {
+      "epoch": 3.9203187250996017,
+      "grad_norm": 0.42267730832099915,
+      "learning_rate": 5.053892906548811e-07,
+      "loss": 0.8960947394371033,
+      "step": 984
+    },
+    {
+      "epoch": 3.9282868525896415,
+      "grad_norm": 1.4257805347442627,
+      "learning_rate": 5.044119612031576e-07,
+      "loss": 0.1468951553106308,
+      "step": 986
+    },
+    {
+      "epoch": 3.9362549800796813,
+      "grad_norm": 1.6065291166305542,
+      "learning_rate": 5.035322435800248e-07,
+      "loss": 0.611916184425354,
+      "step": 988
+    },
+    {
+      "epoch": 3.944223107569721,
+      "grad_norm": 0.1283290535211563,
+      "learning_rate": 5.027501760252857e-07,
+      "loss": 0.577924370765686,
+      "step": 990
+    },
+    {
+      "epoch": 3.952191235059761,
+      "grad_norm": 0.4995441436767578,
+      "learning_rate": 5.02065792534065e-07,
+      "loss": 0.9942194223403931,
+      "step": 992
+    },
+    {
+      "epoch": 3.960159362549801,
+      "grad_norm": 0.5510733723640442,
+      "learning_rate": 5.014791228553294e-07,
+      "loss": 0.3045363128185272,
+      "step": 994
+    },
+    {
+      "epoch": 3.9681274900398407,
+      "grad_norm": 1.766745686531067,
+      "learning_rate": 5.00990192490595e-07,
+      "loss": 0.8708837628364563,
+      "step": 996
+    },
+    {
+      "epoch": 3.9760956175298805,
+      "grad_norm": 0.398499459028244,
+      "learning_rate": 5.005990226928196e-07,
+      "loss": 0.5656142234802246,
+      "step": 998
+    },
+    {
+      "epoch": 3.9840637450199203,
+      "grad_norm": 1.5753365755081177,
+      "learning_rate": 5.003056304654774e-07,
+      "loss": 0.4110683798789978,
+      "step": 1000
+    },
+    {
+      "epoch": 3.99203187250996,
+      "grad_norm": 0.4048484265804291,
+      "learning_rate": 5.001100285618214e-07,
+      "loss": 0.9675297737121582,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.24079236388206482,
+      "learning_rate": 5.000122254843282e-07,
+      "loss": 0.2960621118545532,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0,
+      "step": 1004,
+      "total_flos": 3.519329211045118e+18,
+      "train_loss": 0.8784391032123708,
+      "train_runtime": 9185.186,
+      "train_samples_per_second": 6.558,
+      "train_steps_per_second": 0.109
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1004,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 99999,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.519329211045118e+18,
+  "train_batch_size": 3,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

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