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", 2e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_muon_lr", 2e-5), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_muon_lr", 4e-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", 2e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_muon_lr", 2e-5), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_muon_lr", 4e-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": 7561.8214712507715,
+    "epoch": 4.0,
+    "total_flos": 3.519329211045118e+18,
+    "train_loss": 0.882850165816536,
+    "train_runtime": 9006.0296,
+    "train_samples_per_second": 6.689,
+    "train_steps_per_second": 0.111
+}

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:bc6a48654b2d7c65e58fa5090f6156b1845906a3a8719385b863cdbaf12457e8
+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": 7561.8214712507715,
+    "epoch": 4.0,
+    "total_flos": 3.519329211045118e+18,
+    "train_loss": 0.882850165816536,
+    "train_runtime": 9006.0296,
+    "train_samples_per_second": 6.689,
+    "train_steps_per_second": 0.111
+}

trainer_state.json

@@ -0,0 +1,3557 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 500,
+  "global_step": 1004,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00796812749003984,
+      "grad_norm": 1.0373650789260864,
+      "learning_rate": 9.803921568627452e-08,
+      "loss": 2.09275484085083,
+      "step": 2
+    },
+    {
+      "epoch": 0.01593625498007968,
+      "grad_norm": 1.0212231874465942,
+      "learning_rate": 2.9411764705882356e-07,
+      "loss": 2.010016679763794,
+      "step": 4
+    },
+    {
+      "epoch": 0.02390438247011952,
+      "grad_norm": 2.9287424087524414,
+      "learning_rate": 4.901960784313725e-07,
+      "loss": 2.1995410919189453,
+      "step": 6
+    },
+    {
+      "epoch": 0.03187250996015936,
+      "grad_norm": 0.6216961145401001,
+      "learning_rate": 6.862745098039217e-07,
+      "loss": 1.9172340631484985,
+      "step": 8
+    },
+    {
+      "epoch": 0.0398406374501992,
+      "grad_norm": 2.2636799812316895,
+      "learning_rate": 8.823529411764707e-07,
+      "loss": 1.5700035095214844,
+      "step": 10
+    },
+    {
+      "epoch": 0.04780876494023904,
+      "grad_norm": 0.9179672598838806,
+      "learning_rate": 1.0784313725490197e-06,
+      "loss": 1.815786600112915,
+      "step": 12
+    },
+    {
+      "epoch": 0.055776892430278883,
+      "grad_norm": 0.8624208569526672,
+      "learning_rate": 1.2745098039215686e-06,
+      "loss": 2.7324628829956055,
+      "step": 14
+    },
+    {
+      "epoch": 0.06374501992031872,
+      "grad_norm": 2.463026762008667,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 4.0716986656188965,
+      "step": 16
+    },
+    {
+      "epoch": 0.07171314741035857,
+      "grad_norm": 1.0971407890319824,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.7447350025177002,
+      "step": 18
+    },
+    {
+      "epoch": 0.0796812749003984,
+      "grad_norm": 0.5946111083030701,
+      "learning_rate": 1.8627450980392158e-06,
+      "loss": 1.6588929891586304,
+      "step": 20
+    },
+    {
+      "epoch": 0.08764940239043825,
+      "grad_norm": 0.4733048975467682,
+      "learning_rate": 2.058823529411765e-06,
+      "loss": 1.4994564056396484,
+      "step": 22
+    },
+    {
+      "epoch": 0.09561752988047809,
+      "grad_norm": 1.2157363891601562,
+      "learning_rate": 2.254901960784314e-06,
+      "loss": 1.444083571434021,
+      "step": 24
+    },
+    {
+      "epoch": 0.10358565737051793,
+      "grad_norm": 0.4212270677089691,
+      "learning_rate": 2.450980392156863e-06,
+      "loss": 1.5485053062438965,
+      "step": 26
+    },
+    {
+      "epoch": 0.11155378486055777,
+      "grad_norm": 0.664892315864563,
+      "learning_rate": 2.647058823529412e-06,
+      "loss": 1.6475242376327515,
+      "step": 28
+    },
+    {
+      "epoch": 0.11952191235059761,
+      "grad_norm": 0.9617865681648254,
+      "learning_rate": 2.843137254901961e-06,
+      "loss": 1.4030020236968994,
+      "step": 30
+    },
+    {
+      "epoch": 0.12749003984063745,
+      "grad_norm": 2.0483548641204834,
+      "learning_rate": 3.03921568627451e-06,
+      "loss": 1.2169922590255737,
+      "step": 32
+    },
+    {
+      "epoch": 0.13545816733067728,
+      "grad_norm": 1.2672042846679688,
+      "learning_rate": 3.2352941176470594e-06,
+      "loss": 1.3124995231628418,
+      "step": 34
+    },
+    {
+      "epoch": 0.14342629482071714,
+      "grad_norm": 0.6503031253814697,
+      "learning_rate": 3.431372549019608e-06,
+      "loss": 1.468543529510498,
+      "step": 36
+    },
+    {
+      "epoch": 0.15139442231075698,
+      "grad_norm": 0.5543930530548096,
+      "learning_rate": 3.6274509803921573e-06,
+      "loss": 1.2431751489639282,
+      "step": 38
+    },
+    {
+      "epoch": 0.1593625498007968,
+      "grad_norm": 0.4712694585323334,
+      "learning_rate": 3.8235294117647055e-06,
+      "loss": 1.3735991716384888,
+      "step": 40
+    },
+    {
+      "epoch": 0.16733067729083664,
+      "grad_norm": 0.7119806408882141,
+      "learning_rate": 4.019607843137255e-06,
+      "loss": 0.8355662822723389,
+      "step": 42
+    },
+    {
+      "epoch": 0.1752988047808765,
+      "grad_norm": 0.4997430443763733,
+      "learning_rate": 4.215686274509805e-06,
+      "loss": 1.045559048652649,
+      "step": 44
+    },
+    {
+      "epoch": 0.18326693227091634,
+      "grad_norm": 0.5652483701705933,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 0.9665610194206238,
+      "step": 46
+    },
+    {
+      "epoch": 0.19123505976095617,
+      "grad_norm": 0.6778895258903503,
+      "learning_rate": 4.607843137254902e-06,
+      "loss": 1.2200498580932617,
+      "step": 48
+    },
+    {
+      "epoch": 0.199203187250996,
+      "grad_norm": 0.5004005432128906,
+      "learning_rate": 4.803921568627452e-06,
+      "loss": 1.373177170753479,
+      "step": 50
+    },
+    {
+      "epoch": 0.20717131474103587,
+      "grad_norm": 0.38080382347106934,
+      "learning_rate": 5e-06,
+      "loss": 1.3216673135757446,
+      "step": 52
+    },
+    {
+      "epoch": 0.2151394422310757,
+      "grad_norm": 1.979664921760559,
+      "learning_rate": 4.999951098195543e-06,
+      "loss": 1.1264088153839111,
+      "step": 54
+    },
+    {
+      "epoch": 0.22310756972111553,
+      "grad_norm": 0.5976856350898743,
+      "learning_rate": 4.9998043949078495e-06,
+      "loss": 1.5580120086669922,
+      "step": 56
+    },
+    {
+      "epoch": 0.23107569721115537,
+      "grad_norm": 0.39539989829063416,
+      "learning_rate": 4.9995598965138565e-06,
+      "loss": 1.372806429862976,
+      "step": 58
+    },
+    {
+      "epoch": 0.23904382470119523,
+      "grad_norm": 0.5317901968955994,
+      "learning_rate": 4.9992176136414875e-06,
+      "loss": 0.7824116349220276,
+      "step": 60
+    },
+    {
+      "epoch": 0.24701195219123506,
+      "grad_norm": 0.4483749568462372,
+      "learning_rate": 4.998777561169186e-06,
+      "loss": 1.3768311738967896,
+      "step": 62
+    },
+    {
+      "epoch": 0.2549800796812749,
+      "grad_norm": 0.8371922969818115,
+      "learning_rate": 4.998239758225272e-06,
+      "loss": 1.1183046102523804,
+      "step": 64
+    },
+    {
+      "epoch": 0.26294820717131473,
+      "grad_norm": 0.3723125457763672,
+      "learning_rate": 4.99760422818711e-06,
+      "loss": 1.2992984056472778,
+      "step": 66
+    },
+    {
+      "epoch": 0.27091633466135456,
+      "grad_norm": 0.7791861891746521,
+      "learning_rate": 4.9968709986800925e-06,
+      "loss": 0.803991973400116,
+      "step": 68
+    },
+    {
+      "epoch": 0.2788844621513944,
+      "grad_norm": 1.1810647249221802,
+      "learning_rate": 4.996040101576436e-06,
+      "loss": 1.554394245147705,
+      "step": 70
+    },
+    {
+      "epoch": 0.2868525896414343,
+      "grad_norm": 2.115466594696045,
+      "learning_rate": 4.995111572993804e-06,
+      "loss": 0.8158209323883057,
+      "step": 72
+    },
+    {
+      "epoch": 0.2948207171314741,
+      "grad_norm": 0.9388288259506226,
+      "learning_rate": 4.994085453293725e-06,
+      "loss": 1.135393500328064,
+      "step": 74
+    },
+    {
+      "epoch": 0.30278884462151395,
+      "grad_norm": 0.4037511944770813,
+      "learning_rate": 4.992961787079848e-06,
+      "loss": 1.4617630243301392,
+      "step": 76
+    },
+    {
+      "epoch": 0.3107569721115538,
+      "grad_norm": 1.7258327007293701,
+      "learning_rate": 4.991740623196001e-06,
+      "loss": 0.8883451223373413,
+      "step": 78
+    },
+    {
+      "epoch": 0.3187250996015936,
+      "grad_norm": 0.4252491295337677,
+      "learning_rate": 4.990422014724068e-06,
+      "loss": 0.8365456461906433,
+      "step": 80
+    },
+    {
+      "epoch": 0.32669322709163345,
+      "grad_norm": 1.2421759366989136,
+      "learning_rate": 4.989006018981675e-06,
+      "loss": 1.3236674070358276,
+      "step": 82
+    },
+    {
+      "epoch": 0.3346613545816733,
+      "grad_norm": 0.8827378153800964,
+      "learning_rate": 4.987492697519712e-06,
+      "loss": 0.8737583756446838,
+      "step": 84
+    },
+    {
+      "epoch": 0.3426294820717131,
+      "grad_norm": 0.6238313913345337,
+      "learning_rate": 4.985882116119643e-06,
+      "loss": 1.3124425411224365,
+      "step": 86
+    },
+    {
+      "epoch": 0.350597609561753,
+      "grad_norm": 0.7257739901542664,
+      "learning_rate": 4.9841743447906576e-06,
+      "loss": 0.8785864114761353,
+      "step": 88
+    },
+    {
+      "epoch": 0.35856573705179284,
+      "grad_norm": 0.4855271577835083,
+      "learning_rate": 4.982369457766621e-06,
+      "loss": 1.3321518898010254,
+      "step": 90
+    },
+    {
+      "epoch": 0.3665338645418327,
+      "grad_norm": 1.020969033241272,
+      "learning_rate": 4.980467533502853e-06,
+      "loss": 1.1244926452636719,
+      "step": 92
+    },
+    {
+      "epoch": 0.3745019920318725,
+      "grad_norm": 0.3411503732204437,
+      "learning_rate": 4.978468654672709e-06,
+      "loss": 1.244296669960022,
+      "step": 94
+    },
+    {
+      "epoch": 0.38247011952191234,
+      "grad_norm": 0.9833049178123474,
+      "learning_rate": 4.976372908163999e-06,
+      "loss": 1.098150372505188,
+      "step": 96
+    },
+    {
+      "epoch": 0.3904382470119522,
+      "grad_norm": 0.3009733557701111,
+      "learning_rate": 4.974180385075195e-06,
+      "loss": 1.2701749801635742,
+      "step": 98
+    },
+    {
+      "epoch": 0.398406374501992,
+      "grad_norm": 0.3944694995880127,
+      "learning_rate": 4.971891180711488e-06,
+      "loss": 0.8591183423995972,
+      "step": 100
+    },
+    {
+      "epoch": 0.4063745019920319,
+      "grad_norm": 0.6852508783340454,
+      "learning_rate": 4.9695053945806295e-06,
+      "loss": 1.2809540033340454,
+      "step": 102
+    },
+    {
+      "epoch": 0.41434262948207173,
+      "grad_norm": 0.3390493094921112,
+      "learning_rate": 4.9670231303886194e-06,
+      "loss": 0.8004913926124573,
+      "step": 104
+    },
+    {
+      "epoch": 0.42231075697211157,
+      "grad_norm": 0.8854242563247681,
+      "learning_rate": 4.9644444960351865e-06,
+      "loss": 0.7290127277374268,
+      "step": 106
+    },
+    {
+      "epoch": 0.4302788844621514,
+      "grad_norm": 0.35893553495407104,
+      "learning_rate": 4.9617696036091095e-06,
+      "loss": 0.9962456822395325,
+      "step": 108
+    },
+    {
+      "epoch": 0.43824701195219123,
+      "grad_norm": 0.5066023468971252,
+      "learning_rate": 4.958998569383335e-06,
+      "loss": 1.2905939817428589,
+      "step": 110
+    },
+    {
+      "epoch": 0.44621513944223107,
+      "grad_norm": 0.530055820941925,
+      "learning_rate": 4.956131513809928e-06,
+      "loss": 0.8132796287536621,
+      "step": 112
+    },
+    {
+      "epoch": 0.4541832669322709,
+      "grad_norm": 0.6161649227142334,
+      "learning_rate": 4.953168561514838e-06,
+      "loss": 1.1261539459228516,
+      "step": 114
+    },
+    {
+      "epoch": 0.46215139442231074,
+      "grad_norm": 0.538398802280426,
+      "learning_rate": 4.950109841292475e-06,
+      "loss": 1.2914621829986572,
+      "step": 116
+    },
+    {
+      "epoch": 0.4701195219123506,
+      "grad_norm": 0.45173951983451843,
+      "learning_rate": 4.946955486100119e-06,
+      "loss": 1.3532880544662476,
+      "step": 118
+    },
+    {
+      "epoch": 0.47808764940239046,
+      "grad_norm": 0.3672339916229248,
+      "learning_rate": 4.943705633052134e-06,
+      "loss": 1.237029790878296,
+      "step": 120
+    },
+    {
+      "epoch": 0.4860557768924303,
+      "grad_norm": 0.482098788022995,
+      "learning_rate": 4.940360423414013e-06,
+      "loss": 0.9574871063232422,
+      "step": 122
+    },
+    {
+      "epoch": 0.4940239043824701,
+      "grad_norm": 0.6904265880584717,
+      "learning_rate": 4.936920002596233e-06,
+      "loss": 1.0249451398849487,
+      "step": 124
+    },
+    {
+      "epoch": 0.50199203187251,
+      "grad_norm": 0.8648642897605896,
+      "learning_rate": 4.9333845201479375e-06,
+      "loss": 0.9134154915809631,
+      "step": 126
+    },
+    {
+      "epoch": 0.5099601593625498,
+      "grad_norm": 0.5348114967346191,
+      "learning_rate": 4.9297541297504345e-06,
+      "loss": 1.2641477584838867,
+      "step": 128
+    },
+    {
+      "epoch": 0.5179282868525896,
+      "grad_norm": 0.47338879108428955,
+      "learning_rate": 4.9260289892105164e-06,
+      "loss": 0.9320327639579773,
+      "step": 130
+    },
+    {
+      "epoch": 0.5258964143426295,
+      "grad_norm": 0.48732197284698486,
+      "learning_rate": 4.922209260453599e-06,
+      "loss": 0.8269007205963135,
+      "step": 132
+    },
+    {
+      "epoch": 0.5338645418326693,
+      "grad_norm": 0.5590890049934387,
+      "learning_rate": 4.9182951095166875e-06,
+      "loss": 0.8961949944496155,
+      "step": 134
+    },
+    {
+      "epoch": 0.5418326693227091,
+      "grad_norm": 0.5173343420028687,
+      "learning_rate": 4.914286706541153e-06,
+      "loss": 1.2667150497436523,
+      "step": 136
+    },
+    {
+      "epoch": 0.549800796812749,
+      "grad_norm": 0.5057466626167297,
+      "learning_rate": 4.910184225765341e-06,
+      "loss": 0.9307501316070557,
+      "step": 138
+    },
+    {
+      "epoch": 0.5577689243027888,
+      "grad_norm": 1.0385507345199585,
+      "learning_rate": 4.905987845516996e-06,
+      "loss": 1.0727472305297852,
+      "step": 140
+    },
+    {
+      "epoch": 0.5657370517928287,
+      "grad_norm": 0.5208207964897156,
+      "learning_rate": 4.90169774820551e-06,
+      "loss": 1.0064373016357422,
+      "step": 142
+    },
+    {
+      "epoch": 0.5737051792828686,
+      "grad_norm": 0.497667521238327,
+      "learning_rate": 4.897314120313996e-06,
+      "loss": 1.2229580879211426,
+      "step": 144
+    },
+    {
+      "epoch": 0.5816733067729084,
+      "grad_norm": 0.7592273950576782,
+      "learning_rate": 4.892837152391174e-06,
+      "loss": 1.3348619937896729,
+      "step": 146
+    },
+    {
+      "epoch": 0.5896414342629482,
+      "grad_norm": 0.6640617251396179,
+      "learning_rate": 4.888267039043101e-06,
+      "loss": 0.9323918223381042,
+      "step": 148
+    },
+    {
+      "epoch": 0.5976095617529881,
+      "grad_norm": 1.2756648063659668,
+      "learning_rate": 4.883603978924698e-06,
+      "loss": 0.9315305352210999,
+      "step": 150
+    },
+    {
+      "epoch": 0.6055776892430279,
+      "grad_norm": 1.5417307615280151,
+      "learning_rate": 4.878848174731126e-06,
+      "loss": 0.9146319031715393,
+      "step": 152
+    },
+    {
+      "epoch": 0.6135458167330677,
+      "grad_norm": 0.9504960775375366,
+      "learning_rate": 4.873999833188967e-06,
+      "loss": 1.0927125215530396,
+      "step": 154
+    },
+    {
+      "epoch": 0.6215139442231076,
+      "grad_norm": 4.153389930725098,
+      "learning_rate": 4.8690591650472455e-06,
+      "loss": 0.9705387353897095,
+      "step": 156
+    },
+    {
+      "epoch": 0.6294820717131474,
+      "grad_norm": 1.1229181289672852,
+      "learning_rate": 4.864026385068259e-06,
+      "loss": 1.128150463104248,
+      "step": 158
+    },
+    {
+      "epoch": 0.6374501992031872,
+      "grad_norm": 0.5125619769096375,
+      "learning_rate": 4.858901712018251e-06,
+      "loss": 1.0012102127075195,
+      "step": 160
+    },
+    {
+      "epoch": 0.6454183266932271,
+      "grad_norm": 0.6830053329467773,
+      "learning_rate": 4.8536853686578965e-06,
+      "loss": 0.7507449388504028,
+      "step": 162
+    },
+    {
+      "epoch": 0.6533864541832669,
+      "grad_norm": 0.4223005175590515,
+      "learning_rate": 4.84837758173262e-06,
+      "loss": 1.127695918083191,
+      "step": 164
+    },
+    {
+      "epoch": 0.6613545816733067,
+      "grad_norm": 0.5705804824829102,
+      "learning_rate": 4.8429785819627395e-06,
+      "loss": 0.8070428371429443,
+      "step": 166
+    },
+    {
+      "epoch": 0.6693227091633466,
+      "grad_norm": 0.41760537028312683,
+      "learning_rate": 4.83748860403344e-06,
+      "loss": 1.228546142578125,
+      "step": 168
+    },
+    {
+      "epoch": 0.6772908366533864,
+      "grad_norm": 0.98177570104599,
+      "learning_rate": 4.831907886584566e-06,
+      "loss": 0.891896665096283,
+      "step": 170
+    },
+    {
+      "epoch": 0.6852589641434262,
+      "grad_norm": 0.6809908151626587,
+      "learning_rate": 4.8262366722002535e-06,
+      "loss": 0.8314628601074219,
+      "step": 172
+    },
+    {
+      "epoch": 0.6932270916334662,
+      "grad_norm": 0.5170454382896423,
+      "learning_rate": 4.820475207398386e-06,
+      "loss": 1.2241657972335815,
+      "step": 174
+    },
+    {
+      "epoch": 0.701195219123506,
+      "grad_norm": 2.390455484390259,
+      "learning_rate": 4.8146237426198735e-06,
+      "loss": 0.7440450191497803,
+      "step": 176
+    },
+    {
+      "epoch": 0.7091633466135459,
+      "grad_norm": 1.3662418127059937,
+      "learning_rate": 4.808682532217772e-06,
+      "loss": 0.8962594270706177,
+      "step": 178
+    },
+    {
+      "epoch": 0.7171314741035857,
+      "grad_norm": 6.987403392791748,
+      "learning_rate": 4.802651834446221e-06,
+      "loss": 1.0648951530456543,
+      "step": 180
+    },
+    {
+      "epoch": 0.7250996015936255,
+      "grad_norm": 0.23964297771453857,
+      "learning_rate": 4.796531911449226e-06,
+      "loss": 0.4973796010017395,
+      "step": 182
+    },
+    {
+      "epoch": 0.7330677290836654,
+      "grad_norm": 0.4739725589752197,
+      "learning_rate": 4.790323029249252e-06,
+      "loss": 1.255673885345459,
+      "step": 184
+    },
+    {
+      "epoch": 0.7410358565737052,
+      "grad_norm": 0.5800462961196899,
+      "learning_rate": 4.784025457735675e-06,
+      "loss": 1.240902304649353,
+      "step": 186
+    },
+    {
+      "epoch": 0.749003984063745,
+      "grad_norm": 0.7812846899032593,
+      "learning_rate": 4.777639470653035e-06,
+      "loss": 1.0845593214035034,
+      "step": 188
+    },
+    {
+      "epoch": 0.7569721115537849,
+      "grad_norm": 0.7752678990364075,
+      "learning_rate": 4.771165345589148e-06,
+      "loss": 0.8536431789398193,
+      "step": 190
+    },
+    {
+      "epoch": 0.7649402390438247,
+      "grad_norm": 2.2713398933410645,
+      "learning_rate": 4.7646033639630355e-06,
+      "loss": 1.186915397644043,
+      "step": 192
+    },
+    {
+      "epoch": 0.7729083665338645,
+      "grad_norm": 0.8127148151397705,
+      "learning_rate": 4.757953811012688e-06,
+      "loss": 0.8762093186378479,
+      "step": 194
+    },
+    {
+      "epoch": 0.7808764940239044,
+      "grad_norm": 1.1899384260177612,
+      "learning_rate": 4.7512169757826735e-06,
+      "loss": 0.5605239272117615,
+      "step": 196
+    },
+    {
+      "epoch": 0.7888446215139442,
+      "grad_norm": 1.083921194076538,
+      "learning_rate": 4.74439315111157e-06,
+      "loss": 0.7731846570968628,
+      "step": 198
+    },
+    {
+      "epoch": 0.796812749003984,
+      "grad_norm": 0.44340217113494873,
+      "learning_rate": 4.737482633619234e-06,
+      "loss": 1.0238035917282104,
+      "step": 200
+    },
+    {
+      "epoch": 0.8047808764940239,
+      "grad_norm": 0.2869577705860138,
+      "learning_rate": 4.730485723693912e-06,
+      "loss": 0.9578498005867004,
+      "step": 202
+    },
+    {
+      "epoch": 0.8127490039840638,
+      "grad_norm": 1.9805564880371094,
+      "learning_rate": 4.723402725479177e-06,
+      "loss": 1.4975579977035522,
+      "step": 204
+    },
+    {
+      "epoch": 0.8207171314741036,
+      "grad_norm": 3.5942680835723877,
+      "learning_rate": 4.71623394686071e-06,
+      "loss": 1.2103545665740967,
+      "step": 206
+    },
+    {
+      "epoch": 0.8286852589641435,
+      "grad_norm": 0.3902193307876587,
+      "learning_rate": 4.708979699452923e-06,
+      "loss": 0.5866236090660095,
+      "step": 208
+    },
+    {
+      "epoch": 0.8366533864541833,
+      "grad_norm": 0.5997117757797241,
+      "learning_rate": 4.701640298585405e-06,
+      "loss": 1.253372073173523,
+      "step": 210
+    },
+    {
+      "epoch": 0.8446215139442231,
+      "grad_norm": 0.4733811914920807,
+      "learning_rate": 4.694216063289219e-06,
+      "loss": 1.2103458642959595,
+      "step": 212
+    },
+    {
+      "epoch": 0.852589641434263,
+      "grad_norm": 0.37428876757621765,
+      "learning_rate": 4.686707316283033e-06,
+      "loss": 0.925890326499939,
+      "step": 214
+    },
+    {
+      "epoch": 0.8605577689243028,
+      "grad_norm": 0.8632296323776245,
+      "learning_rate": 4.679114383959095e-06,
+      "loss": 1.4079800844192505,
+      "step": 216
+    },
+    {
+      "epoch": 0.8685258964143426,
+      "grad_norm": 1.0630208253860474,
+      "learning_rate": 4.6714375963690415e-06,
+      "loss": 0.8563957214355469,
+      "step": 218
+    },
+    {
+      "epoch": 0.8764940239043825,
+      "grad_norm": 0.572202205657959,
+      "learning_rate": 4.6636772872095545e-06,
+      "loss": 1.2282161712646484,
+      "step": 220
+    },
+    {
+      "epoch": 0.8844621513944223,
+      "grad_norm": 0.8659475445747375,
+      "learning_rate": 4.6558337938078505e-06,
+      "loss": 1.473631739616394,
+      "step": 222
+    },
+    {
+      "epoch": 0.8924302788844621,
+      "grad_norm": 0.5699397325515747,
+      "learning_rate": 4.647907457107025e-06,
+      "loss": 0.6868866086006165,
+      "step": 224
+    },
+    {
+      "epoch": 0.900398406374502,
+      "grad_norm": 0.7301930785179138,
+      "learning_rate": 4.639898621651226e-06,
+      "loss": 0.9735338091850281,
+      "step": 226
+    },
+    {
+      "epoch": 0.9083665338645418,
+      "grad_norm": 5.144070625305176,
+      "learning_rate": 4.631807635570681e-06,
+      "loss": 1.0671145915985107,
+      "step": 228
+    },
+    {
+      "epoch": 0.9163346613545816,
+      "grad_norm": 0.7398518919944763,
+      "learning_rate": 4.623634850566563e-06,
+      "loss": 1.172678828239441,
+      "step": 230
+    },
+    {
+      "epoch": 0.9243027888446215,
+      "grad_norm": 0.426059365272522,
+      "learning_rate": 4.615380621895701e-06,
+      "loss": 1.3170030117034912,
+      "step": 232
+    },
+    {
+      "epoch": 0.9322709163346613,
+      "grad_norm": 0.5351225733757019,
+      "learning_rate": 4.607045308355141e-06,
+      "loss": 0.7853307723999023,
+      "step": 234
+    },
+    {
+      "epoch": 0.9402390438247012,
+      "grad_norm": 0.5201543569564819,
+      "learning_rate": 4.598629272266549e-06,
+      "loss": 0.8055428862571716,
+      "step": 236
+    },
+    {
+      "epoch": 0.9482071713147411,
+      "grad_norm": 1.9418190717697144,
+      "learning_rate": 4.590132879460458e-06,
+      "loss": 0.4271475374698639,
+      "step": 238
+    },
+    {
+      "epoch": 0.9561752988047809,
+      "grad_norm": 1.7852131128311157,
+      "learning_rate": 4.5815564992603705e-06,
+      "loss": 0.8238165974617004,
+      "step": 240
+    },
+    {
+      "epoch": 0.9641434262948207,
+      "grad_norm": 0.6683778762817383,
+      "learning_rate": 4.572900504466702e-06,
+      "loss": 1.1197185516357422,
+      "step": 242
+    },
+    {
+      "epoch": 0.9721115537848606,
+      "grad_norm": 0.6745935678482056,
+      "learning_rate": 4.564165271340577e-06,
+      "loss": 1.2074919939041138,
+      "step": 244
+    },
+    {
+      "epoch": 0.9800796812749004,
+      "grad_norm": 0.6725598573684692,
+      "learning_rate": 4.555351179587472e-06,
+      "loss": 0.9544854164123535,
+      "step": 246
+    },
+    {
+      "epoch": 0.9880478087649402,
+      "grad_norm": 0.3850212097167969,
+      "learning_rate": 4.546458612340714e-06,
+      "loss": 0.9788272380828857,
+      "step": 248
+    },
+    {
+      "epoch": 0.9960159362549801,
+      "grad_norm": 0.4695571959018707,
+      "learning_rate": 4.537487956144821e-06,
+      "loss": 0.9692898988723755,
+      "step": 250
+    },
+    {
+      "epoch": 1.00398406374502,
+      "grad_norm": 0.4883769750595093,
+      "learning_rate": 4.528439600938706e-06,
+      "loss": 0.9087216258049011,
+      "step": 252
+    },
+    {
+      "epoch": 1.0119521912350598,
+      "grad_norm": 0.33964306116104126,
+      "learning_rate": 4.519313940038721e-06,
+      "loss": 0.6800986528396606,
+      "step": 254
+    },
+    {
+      "epoch": 1.0199203187250996,
+      "grad_norm": 0.46170279383659363,
+      "learning_rate": 4.510111370121563e-06,
+      "loss": 1.1623857021331787,
+      "step": 256
+    },
+    {
+      "epoch": 1.0278884462151394,
+      "grad_norm": 0.5654280781745911,
+      "learning_rate": 4.500832291207033e-06,
+      "loss": 1.1357818841934204,
+      "step": 258
+    },
+    {
+      "epoch": 1.0358565737051793,
+      "grad_norm": 0.517961859703064,
+      "learning_rate": 4.491477106640646e-06,
+      "loss": 1.048874020576477,
+      "step": 260
+    },
+    {
+      "epoch": 1.043824701195219,
+      "grad_norm": 0.5235061645507812,
+      "learning_rate": 4.482046223076092e-06,
+      "loss": 1.016953945159912,
+      "step": 262
+    },
+    {
+      "epoch": 1.051792828685259,
+      "grad_norm": 1.1011329889297485,
+      "learning_rate": 4.4725400504575725e-06,
+      "loss": 1.060608983039856,
+      "step": 264
+    },
+    {
+      "epoch": 1.0597609561752988,
+      "grad_norm": 1.369287371635437,
+      "learning_rate": 4.462959002001971e-06,
+      "loss": 0.5993220806121826,
+      "step": 266
+    },
+    {
+      "epoch": 1.0677290836653386,
+      "grad_norm": 2.090449094772339,
+      "learning_rate": 4.453303494180895e-06,
+      "loss": 0.9326719641685486,
+      "step": 268
+    },
+    {
+      "epoch": 1.0756972111553784,
+      "grad_norm": 0.3588223457336426,
+      "learning_rate": 4.443573946702573e-06,
+      "loss": 0.5480161905288696,
+      "step": 270
+    },
+    {
+      "epoch": 1.0836653386454183,
+      "grad_norm": 0.44564101099967957,
+      "learning_rate": 4.433770782493606e-06,
+      "loss": 1.2314869165420532,
+      "step": 272
+    },
+    {
+      "epoch": 1.091633466135458,
+      "grad_norm": 1.0386971235275269,
+      "learning_rate": 4.42389442768059e-06,
+      "loss": 0.5619557499885559,
+      "step": 274
+    },
+    {
+      "epoch": 1.099601593625498,
+      "grad_norm": 0.44276317954063416,
+      "learning_rate": 4.413945311571586e-06,
+      "loss": 0.9132966995239258,
+      "step": 276
+    },
+    {
+      "epoch": 1.1075697211155378,
+      "grad_norm": 0.39500024914741516,
+      "learning_rate": 4.40392386663747e-06,
+      "loss": 0.8292908668518066,
+      "step": 278
+    },
+    {
+      "epoch": 1.1155378486055776,
+      "grad_norm": 0.7742955088615417,
+      "learning_rate": 4.393830528493118e-06,
+      "loss": 1.0214531421661377,
+      "step": 280
+    },
+    {
+      "epoch": 1.1235059760956174,
+      "grad_norm": 0.7359810471534729,
+      "learning_rate": 4.3836657358784875e-06,
+      "loss": 1.1593756675720215,
+      "step": 282
+    },
+    {
+      "epoch": 1.1314741035856573,
+      "grad_norm": 0.8002145886421204,
+      "learning_rate": 4.373429930639534e-06,
+      "loss": 1.1806137561798096,
+      "step": 284
+    },
+    {
+      "epoch": 1.139442231075697,
+      "grad_norm": 0.31167158484458923,
+      "learning_rate": 4.3631235577090085e-06,
+      "loss": 0.2554236054420471,
+      "step": 286
+    },
+    {
+      "epoch": 1.1474103585657371,
+      "grad_norm": 0.25866594910621643,
+      "learning_rate": 4.35274706508712e-06,
+      "loss": 0.71672123670578,
+      "step": 288
+    },
+    {
+      "epoch": 1.155378486055777,
+      "grad_norm": 0.724770724773407,
+      "learning_rate": 4.342300903822057e-06,
+      "loss": 0.900907039642334,
+      "step": 290
+    },
+    {
+      "epoch": 1.1633466135458168,
+      "grad_norm": 0.706335723400116,
+      "learning_rate": 4.331785527990384e-06,
+      "loss": 0.5006822943687439,
+      "step": 292
+    },
+    {
+      "epoch": 1.1713147410358566,
+      "grad_norm": 0.8979578614234924,
+      "learning_rate": 4.3212013946773045e-06,
+      "loss": 1.216837763786316,
+      "step": 294
+    },
+    {
+      "epoch": 1.1792828685258965,
+      "grad_norm": 0.5903942584991455,
+      "learning_rate": 4.31054896395679e-06,
+      "loss": 1.2181227207183838,
+      "step": 296
+    },
+    {
+      "epoch": 1.1872509960159363,
+      "grad_norm": 0.49522045254707336,
+      "learning_rate": 4.299828698871581e-06,
+      "loss": 1.3492027521133423,
+      "step": 298
+    },
+    {
+      "epoch": 1.1952191235059761,
+      "grad_norm": 42.5111083984375,
+      "learning_rate": 4.289041065413062e-06,
+      "loss": 0.6173719167709351,
+      "step": 300
+    },
+    {
+      "epoch": 1.203187250996016,
+      "grad_norm": 0.7985536456108093,
+      "learning_rate": 4.278186532501003e-06,
+      "loss": 0.9192322492599487,
+      "step": 302
+    },
+    {
+      "epoch": 1.2111553784860558,
+      "grad_norm": 2.0916953086853027,
+      "learning_rate": 4.267265571963179e-06,
+      "loss": 0.5548972487449646,
+      "step": 304
+    },
+    {
+      "epoch": 1.2191235059760956,
+      "grad_norm": 1.433501124382019,
+      "learning_rate": 4.256278658514858e-06,
+      "loss": 1.1697511672973633,
+      "step": 306
+    },
+    {
+      "epoch": 1.2270916334661355,
+      "grad_norm": 2.002721071243286,
+      "learning_rate": 4.245226269738165e-06,
+      "loss": 1.0041544437408447,
+      "step": 308
+    },
+    {
+      "epoch": 1.2350597609561753,
+      "grad_norm": 0.8270197510719299,
+      "learning_rate": 4.23410888606133e-06,
+      "loss": 1.1297266483306885,
+      "step": 310
+    },
+    {
+      "epoch": 1.2430278884462151,
+      "grad_norm": 0.7538009881973267,
+      "learning_rate": 4.222926990737793e-06,
+      "loss": 1.0430384874343872,
+      "step": 312
+    },
+    {
+      "epoch": 1.250996015936255,
+      "grad_norm": 1.0800330638885498,
+      "learning_rate": 4.211681069825207e-06,
+      "loss": 0.8671213388442993,
+      "step": 314
+    },
+    {
+      "epoch": 1.2589641434262948,
+      "grad_norm": 0.858415961265564,
+      "learning_rate": 4.200371612164305e-06,
+      "loss": 0.7203362584114075,
+      "step": 316
+    },
+    {
+      "epoch": 1.2669322709163346,
+      "grad_norm": 0.7451192140579224,
+      "learning_rate": 4.1889991093576555e-06,
+      "loss": 0.6065669655799866,
+      "step": 318
+    },
+    {
+      "epoch": 1.2749003984063745,
+      "grad_norm": 0.7658272981643677,
+      "learning_rate": 4.177564055748284e-06,
+      "loss": 0.5560135841369629,
+      "step": 320
+    },
+    {
+      "epoch": 1.2828685258964143,
+      "grad_norm": 0.7753592729568481,
+      "learning_rate": 4.166066948398199e-06,
+      "loss": 0.9698732495307922,
+      "step": 322
+    },
+    {
+      "epoch": 1.2908366533864541,
+      "grad_norm": 1.1220102310180664,
+      "learning_rate": 4.154508287066772e-06,
+      "loss": 0.9322744607925415,
+      "step": 324
+    },
+    {
+      "epoch": 1.298804780876494,
+      "grad_norm": 3.4268646240234375,
+      "learning_rate": 4.142888574189025e-06,
+      "loss": 1.2757772207260132,
+      "step": 326
+    },
+    {
+      "epoch": 1.3067729083665338,
+      "grad_norm": 0.9928934574127197,
+      "learning_rate": 4.131208314853778e-06,
+      "loss": 0.6775528788566589,
+      "step": 328
+    },
+    {
+      "epoch": 1.3147410358565736,
+      "grad_norm": 0.3143123388290405,
+      "learning_rate": 4.119468016781707e-06,
+      "loss": 0.8935784697532654,
+      "step": 330
+    },
+    {
+      "epoch": 1.3227091633466135,
+      "grad_norm": 0.6986421942710876,
+      "learning_rate": 4.107668190303266e-06,
+      "loss": 0.9664567112922668,
+      "step": 332
+    },
+    {
+      "epoch": 1.3306772908366533,
+      "grad_norm": 2.046541690826416,
+      "learning_rate": 4.095809348336506e-06,
+      "loss": 0.613030195236206,
+      "step": 334
+    },
+    {
+      "epoch": 1.3386454183266931,
+      "grad_norm": 0.6653932929039001,
+      "learning_rate": 4.083892006364777e-06,
+      "loss": 0.8722373843193054,
+      "step": 336
+    },
+    {
+      "epoch": 1.3466135458167332,
+      "grad_norm": 1.290967583656311,
+      "learning_rate": 4.071916682414327e-06,
+      "loss": 1.138928771018982,
+      "step": 338
+    },
+    {
+      "epoch": 1.354581673306773,
+      "grad_norm": 0.5062804818153381,
+      "learning_rate": 4.059883897031777e-06,
+      "loss": 1.1701346635818481,
+      "step": 340
+    },
+    {
+      "epoch": 1.3625498007968129,
+      "grad_norm": 0.6293162703514099,
+      "learning_rate": 4.0477941732615e-06,
+      "loss": 1.1722662448883057,
+      "step": 342
+    },
+    {
+      "epoch": 1.3705179282868527,
+      "grad_norm": 0.854015052318573,
+      "learning_rate": 4.035648036622879e-06,
+      "loss": 1.1671116352081299,
+      "step": 344
+    },
+    {
+      "epoch": 1.3784860557768925,
+      "grad_norm": 0.6483955383300781,
+      "learning_rate": 4.023446015087469e-06,
+      "loss": 0.6964617967605591,
+      "step": 346
+    },
+    {
+      "epoch": 1.3864541832669324,
+      "grad_norm": 0.5509741306304932,
+      "learning_rate": 4.0111886390560446e-06,
+      "loss": 1.1621150970458984,
+      "step": 348
+    },
+    {
+      "epoch": 1.3944223107569722,
+      "grad_norm": 0.5316842198371887,
+      "learning_rate": 3.998876441335544e-06,
+      "loss": 0.6768369674682617,
+      "step": 350
+    },
+    {
+      "epoch": 1.402390438247012,
+      "grad_norm": 0.4520191550254822,
+      "learning_rate": 3.98650995711591e-06,
+      "loss": 1.1876813173294067,
+      "step": 352
+    },
+    {
+      "epoch": 1.4103585657370519,
+      "grad_norm": 0.6878074407577515,
+      "learning_rate": 3.974089723946826e-06,
+      "loss": 1.1137140989303589,
+      "step": 354
+    },
+    {
+      "epoch": 1.4183266932270917,
+      "grad_norm": 1.2717393636703491,
+      "learning_rate": 3.9616162817143465e-06,
+      "loss": 0.9824956059455872,
+      "step": 356
+    },
+    {
+      "epoch": 1.4262948207171315,
+      "grad_norm": 0.449339896440506,
+      "learning_rate": 3.949090172617437e-06,
+      "loss": 1.1469612121582031,
+      "step": 358
+    },
+    {
+      "epoch": 1.4342629482071714,
+      "grad_norm": 1.307947039604187,
+      "learning_rate": 3.936511941144397e-06,
+      "loss": 0.5965569615364075,
+      "step": 360
+    },
+    {
+      "epoch": 1.4422310756972112,
+      "grad_norm": 0.651824414730072,
+      "learning_rate": 3.9238821340491936e-06,
+      "loss": 0.7194271087646484,
+      "step": 362
+    },
+    {
+      "epoch": 1.450199203187251,
+      "grad_norm": 0.923170268535614,
+      "learning_rate": 3.911201300327703e-06,
+      "loss": 1.0449382066726685,
+      "step": 364
+    },
+    {
+      "epoch": 1.4581673306772909,
+      "grad_norm": 0.3838039040565491,
+      "learning_rate": 3.898469991193835e-06,
+      "loss": 1.1666339635849,
+      "step": 366
+    },
+    {
+      "epoch": 1.4661354581673307,
+      "grad_norm": 0.5390955805778503,
+      "learning_rate": 3.885688760055581e-06,
+      "loss": 0.6083441376686096,
+      "step": 368
+    },
+    {
+      "epoch": 1.4741035856573705,
+      "grad_norm": 0.40487271547317505,
+      "learning_rate": 3.872858162490956e-06,
+      "loss": 0.6765357255935669,
+      "step": 370
+    },
+    {
+      "epoch": 1.4820717131474104,
+      "grad_norm": 3.386698007583618,
+      "learning_rate": 3.859978756223845e-06,
+      "loss": 1.21755051612854,
+      "step": 372
+    },
+    {
+      "epoch": 1.4900398406374502,
+      "grad_norm": 0.4938630163669586,
+      "learning_rate": 3.847051101099768e-06,
+      "loss": 1.140271544456482,
+      "step": 374
+    },
+    {
+      "epoch": 1.49800796812749,
+      "grad_norm": 1.8366533517837524,
+      "learning_rate": 3.834075759061535e-06,
+      "loss": 1.1353375911712646,
+      "step": 376
+    },
+    {
+      "epoch": 1.5059760956175299,
+      "grad_norm": 1.8885377645492554,
+      "learning_rate": 3.8210532941248255e-06,
+      "loss": 0.907731294631958,
+      "step": 378
+    },
+    {
+      "epoch": 1.5139442231075697,
+      "grad_norm": 0.44331249594688416,
+      "learning_rate": 3.8079842723536688e-06,
+      "loss": 1.134822130203247,
+      "step": 380
+    },
+    {
+      "epoch": 1.5219123505976095,
+      "grad_norm": 0.4130204916000366,
+      "learning_rate": 3.7948692618358424e-06,
+      "loss": 0.7457793354988098,
+      "step": 382
+    },
+    {
+      "epoch": 1.5298804780876494,
+      "grad_norm": 0.4773809015750885,
+      "learning_rate": 3.781708832658169e-06,
+      "loss": 0.748079776763916,
+      "step": 384
+    },
+    {
+      "epoch": 1.5378486055776892,
+      "grad_norm": 1.184057593345642,
+      "learning_rate": 3.7685035568817492e-06,
+      "loss": 1.0133119821548462,
+      "step": 386
+    },
+    {
+      "epoch": 1.545816733067729,
+      "grad_norm": 0.7657914161682129,
+      "learning_rate": 3.7552540085170826e-06,
+      "loss": 1.1418333053588867,
+      "step": 388
+    },
+    {
+      "epoch": 1.5537848605577689,
+      "grad_norm": 1.495758295059204,
+      "learning_rate": 3.7419607634991235e-06,
+      "loss": 0.5878637433052063,
+      "step": 390
+    },
+    {
+      "epoch": 1.5617529880478087,
+      "grad_norm": 0.8114638328552246,
+      "learning_rate": 3.7286243996622432e-06,
+      "loss": 0.5885580778121948,
+      "step": 392
+    },
+    {
+      "epoch": 1.5697211155378485,
+      "grad_norm": 1.4311312437057495,
+      "learning_rate": 3.715245496715116e-06,
+      "loss": 1.1066184043884277,
+      "step": 394
+    },
+    {
+      "epoch": 1.5776892430278884,
+      "grad_norm": 1.2078579664230347,
+      "learning_rate": 3.7018246362155134e-06,
+      "loss": 1.0669033527374268,
+      "step": 396
+    },
+    {
+      "epoch": 1.5856573705179282,
+      "grad_norm": 0.3787723481655121,
+      "learning_rate": 3.6883624015450344e-06,
+      "loss": 1.134636640548706,
+      "step": 398
+    },
+    {
+      "epoch": 1.593625498007968,
+      "grad_norm": 0.7368590235710144,
+      "learning_rate": 3.6748593778837367e-06,
+      "loss": 0.6034888029098511,
+      "step": 400
+    },
+    {
+      "epoch": 1.6015936254980079,
+      "grad_norm": 0.913153350353241,
+      "learning_rate": 3.661316152184708e-06,
+      "loss": 1.116573452949524,
+      "step": 402
+    },
+    {
+      "epoch": 1.6095617529880477,
+      "grad_norm": 1.0158803462982178,
+      "learning_rate": 3.6477333131485483e-06,
+      "loss": 1.263643741607666,
+      "step": 404
+    },
+    {
+      "epoch": 1.6175298804780875,
+      "grad_norm": 1.1920008659362793,
+      "learning_rate": 3.634111451197781e-06,
+      "loss": 0.7203158736228943,
+      "step": 406
+    },
+    {
+      "epoch": 1.6254980079681274,
+      "grad_norm": 0.3185639977455139,
+      "learning_rate": 3.6204511584511865e-06,
+      "loss": 0.5686110258102417,
+      "step": 408
+    },
+    {
+      "epoch": 1.6334661354581672,
+      "grad_norm": 0.32994651794433594,
+      "learning_rate": 3.606753028698067e-06,
+      "loss": 0.7672407627105713,
+      "step": 410
+    },
+    {
+      "epoch": 1.641434262948207,
+      "grad_norm": 1.2779107093811035,
+      "learning_rate": 3.593017657372436e-06,
+      "loss": 0.5063269734382629,
+      "step": 412
+    },
+    {
+      "epoch": 1.6494023904382469,
+      "grad_norm": 0.2983052432537079,
+      "learning_rate": 3.5792456415271288e-06,
+      "loss": 0.8370915055274963,
+      "step": 414
+    },
+    {
+      "epoch": 1.6573705179282867,
+      "grad_norm": 1.2749823331832886,
+      "learning_rate": 3.565437579807858e-06,
+      "loss": 0.7320247888565063,
+      "step": 416
+    },
+    {
+      "epoch": 1.6653386454183265,
+      "grad_norm": 3.201230049133301,
+      "learning_rate": 3.551594072427188e-06,
+      "loss": 0.9960025548934937,
+      "step": 418
+    },
+    {
+      "epoch": 1.6733067729083664,
+      "grad_norm": 0.7669857740402222,
+      "learning_rate": 3.537715721138444e-06,
+      "loss": 1.1347576379776,
+      "step": 420
+    },
+    {
+      "epoch": 1.6812749003984062,
+      "grad_norm": 0.39680543541908264,
+      "learning_rate": 3.523803129209554e-06,
+      "loss": 1.01784348487854,
+      "step": 422
+    },
+    {
+      "epoch": 1.6892430278884463,
+      "grad_norm": 0.6956140398979187,
+      "learning_rate": 3.5098569013968314e-06,
+      "loss": 0.5958586931228638,
+      "step": 424
+    },
+    {
+      "epoch": 1.697211155378486,
+      "grad_norm": 0.4441296458244324,
+      "learning_rate": 3.49587764391868e-06,
+      "loss": 1.2017691135406494,
+      "step": 426
+    },
+    {
+      "epoch": 1.705179282868526,
+      "grad_norm": 0.33807823061943054,
+      "learning_rate": 3.4818659644292473e-06,
+      "loss": 1.1366020441055298,
+      "step": 428
+    },
+    {
+      "epoch": 1.7131474103585658,
+      "grad_norm": 0.7394194006919861,
+      "learning_rate": 3.4678224719920085e-06,
+      "loss": 0.7497201561927795,
+      "step": 430
+    },
+    {
+      "epoch": 1.7211155378486056,
+      "grad_norm": 0.7117893695831299,
+      "learning_rate": 3.4537477770532933e-06,
+      "loss": 0.9368507266044617,
+      "step": 432
+    },
+    {
+      "epoch": 1.7290836653386454,
+      "grad_norm": 0.8230752348899841,
+      "learning_rate": 3.4396424914157527e-06,
+      "loss": 0.6528562307357788,
+      "step": 434
+    },
+    {
+      "epoch": 1.7370517928286853,
+      "grad_norm": 2.007270574569702,
+      "learning_rate": 3.4255072282117574e-06,
+      "loss": 0.3780575394630432,
+      "step": 436
+    },
+    {
+      "epoch": 1.745019920318725,
+      "grad_norm": 0.4594070613384247,
+      "learning_rate": 3.411342601876754e-06,
+      "loss": 0.7573771476745605,
+      "step": 438
+    },
+    {
+      "epoch": 1.752988047808765,
+      "grad_norm": 0.9193369150161743,
+      "learning_rate": 3.3971492281225554e-06,
+      "loss": 0.6597039699554443,
+      "step": 440
+    },
+    {
+      "epoch": 1.7609561752988048,
+      "grad_norm": 0.8584269285202026,
+      "learning_rate": 3.382927723910572e-06,
+      "loss": 1.600492238998413,
+      "step": 442
+    },
+    {
+      "epoch": 1.7689243027888446,
+      "grad_norm": 2.620512008666992,
+      "learning_rate": 3.3686787074249966e-06,
+      "loss": 0.7913353443145752,
+      "step": 444
+    },
+    {
+      "epoch": 1.7768924302788844,
+      "grad_norm": 0.49178755283355713,
+      "learning_rate": 3.354402798045935e-06,
+      "loss": 1.0425364971160889,
+      "step": 446
+    },
+    {
+      "epoch": 1.7848605577689243,
+      "grad_norm": 0.7130312323570251,
+      "learning_rate": 3.340100616322479e-06,
+      "loss": 1.0970323085784912,
+      "step": 448
+    },
+    {
+      "epoch": 1.792828685258964,
+      "grad_norm": 0.3831311762332916,
+      "learning_rate": 3.3257727839457356e-06,
+      "loss": 0.8711647987365723,
+      "step": 450
+    },
+    {
+      "epoch": 1.800796812749004,
+      "grad_norm": 0.4964040517807007,
+      "learning_rate": 3.311419923721798e-06,
+      "loss": 0.5358787775039673,
+      "step": 452
+    },
+    {
+      "epoch": 1.8087649402390438,
+      "grad_norm": 0.6216398477554321,
+      "learning_rate": 3.29704265954468e-06,
+      "loss": 0.9406894445419312,
+      "step": 454
+    },
+    {
+      "epoch": 1.8167330677290838,
+      "grad_norm": 1.3005354404449463,
+      "learning_rate": 3.282641616369193e-06,
+      "loss": 0.667259931564331,
+      "step": 456
+    },
+    {
+      "epoch": 1.8247011952191237,
+      "grad_norm": 1.1243559122085571,
+      "learning_rate": 3.2682174201837773e-06,
+      "loss": 1.0456857681274414,
+      "step": 458
+    },
+    {
+      "epoch": 1.8326693227091635,
+      "grad_norm": 2.3342390060424805,
+      "learning_rate": 3.2537706979832955e-06,
+      "loss": 0.7288594245910645,
+      "step": 460
+    },
+    {
+      "epoch": 1.8406374501992033,
+      "grad_norm": 0.4283381998538971,
+      "learning_rate": 3.23930207774178e-06,
+      "loss": 0.7739078402519226,
+      "step": 462
+    },
+    {
+      "epoch": 1.8486055776892432,
+      "grad_norm": 0.21832984685897827,
+      "learning_rate": 3.2248121883851296e-06,
+      "loss": 0.11355259269475937,
+      "step": 464
+    },
+    {
+      "epoch": 1.856573705179283,
+      "grad_norm": 0.5141590237617493,
+      "learning_rate": 3.210301659763777e-06,
+      "loss": 1.1529980897903442,
+      "step": 466
+    },
+    {
+      "epoch": 1.8645418326693228,
+      "grad_norm": 5.826035022735596,
+      "learning_rate": 3.1957711226253073e-06,
+      "loss": 0.5443772077560425,
+      "step": 468
+    },
+    {
+      "epoch": 1.8725099601593627,
+      "grad_norm": 4.228695392608643,
+      "learning_rate": 3.181221208587041e-06,
+      "loss": 1.1379239559173584,
+      "step": 470
+    },
+    {
+      "epoch": 1.8804780876494025,
+      "grad_norm": 0.5911218523979187,
+      "learning_rate": 3.16665255010858e-06,
+      "loss": 1.0379854440689087,
+      "step": 472
+    },
+    {
+      "epoch": 1.8884462151394423,
+      "grad_norm": 1.448185920715332,
+      "learning_rate": 3.152065780464311e-06,
+      "loss": 1.1778509616851807,
+      "step": 474
+    },
+    {
+      "epoch": 1.8964143426294822,
+      "grad_norm": 0.7432910799980164,
+      "learning_rate": 3.1374615337158874e-06,
+      "loss": 0.9662929177284241,
+      "step": 476
+    },
+    {
+      "epoch": 1.904382470119522,
+      "grad_norm": 0.48240557312965393,
+      "learning_rate": 3.1228404446846577e-06,
+      "loss": 0.8983007073402405,
+      "step": 478
+    },
+    {
+      "epoch": 1.9123505976095618,
+      "grad_norm": 0.7894963622093201,
+      "learning_rate": 3.1082031489240775e-06,
+      "loss": 0.5984846949577332,
+      "step": 480
+    },
+    {
+      "epoch": 1.9203187250996017,
+      "grad_norm": 0.7460091710090637,
+      "learning_rate": 3.093550282692081e-06,
+      "loss": 0.7599999904632568,
+      "step": 482
+    },
+    {
+      "epoch": 1.9282868525896415,
+      "grad_norm": 0.4634132981300354,
+      "learning_rate": 3.0788824829234206e-06,
+      "loss": 1.1132484674453735,
+      "step": 484
+    },
+    {
+      "epoch": 1.9362549800796813,
+      "grad_norm": 0.6595823764801025,
+      "learning_rate": 3.064200387201987e-06,
+      "loss": 1.1532316207885742,
+      "step": 486
+    },
+    {
+      "epoch": 1.9442231075697212,
+      "grad_norm": 1.4694468975067139,
+      "learning_rate": 3.0495046337330917e-06,
+      "loss": 1.1064248085021973,
+      "step": 488
+    },
+    {
+      "epoch": 1.952191235059761,
+      "grad_norm": 0.3540489673614502,
+      "learning_rate": 3.034795861315723e-06,
+      "loss": 1.11617112159729,
+      "step": 490
+    },
+    {
+      "epoch": 1.9601593625498008,
+      "grad_norm": 0.6252588033676147,
+      "learning_rate": 3.020074709314781e-06,
+      "loss": 0.7896658182144165,
+      "step": 492
+    },
+    {
+      "epoch": 1.9681274900398407,
+      "grad_norm": 6.2591447830200195,
+      "learning_rate": 3.0053418176332844e-06,
+      "loss": 0.6371893882751465,
+      "step": 494
+    },
+    {
+      "epoch": 1.9760956175298805,
+      "grad_norm": 0.3901083469390869,
+      "learning_rate": 2.9905978266845565e-06,
+      "loss": 1.0997235774993896,
+      "step": 496
+    },
+    {
+      "epoch": 1.9840637450199203,
+      "grad_norm": 0.8827555775642395,
+      "learning_rate": 2.975843377364385e-06,
+      "loss": 0.8393293619155884,
+      "step": 498
+    },
+    {
+      "epoch": 1.9920318725099602,
+      "grad_norm": 0.5440815687179565,
+      "learning_rate": 2.961079111023167e-06,
+      "loss": 1.1662544012069702,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6568570137023926,
+      "learning_rate": 2.9463056694380263e-06,
+      "loss": 0.6173162460327148,
+      "step": 502
+    },
+    {
+      "epoch": 2.00796812749004,
+      "grad_norm": 0.39056989550590515,
+      "learning_rate": 2.9315236947849213e-06,
+      "loss": 0.6433302164077759,
+      "step": 504
+    },
+    {
+      "epoch": 2.0159362549800797,
+      "grad_norm": 0.7750610709190369,
+      "learning_rate": 2.916733829610726e-06,
+      "loss": 0.5944309234619141,
+      "step": 506
+    },
+    {
+      "epoch": 2.0239043824701195,
+      "grad_norm": 0.3872879147529602,
+      "learning_rate": 2.901936716805302e-06,
+      "loss": 0.5495409369468689,
+      "step": 508
+    },
+    {
+      "epoch": 2.0318725099601593,
+      "grad_norm": 0.8868218660354614,
+      "learning_rate": 2.887132999573554e-06,
+      "loss": 0.9489161968231201,
+      "step": 510
+    },
+    {
+      "epoch": 2.039840637450199,
+      "grad_norm": 0.7914445400238037,
+      "learning_rate": 2.8723233214074675e-06,
+      "loss": 0.7698367238044739,
+      "step": 512
+    },
+    {
+      "epoch": 2.047808764940239,
+      "grad_norm": 33.907936096191406,
+      "learning_rate": 2.857508326058142e-06,
+      "loss": 0.28293168544769287,
+      "step": 514
+    },
+    {
+      "epoch": 2.055776892430279,
+      "grad_norm": 0.42631280422210693,
+      "learning_rate": 2.842688657507805e-06,
+      "loss": 0.9104582071304321,
+      "step": 516
+    },
+    {
+      "epoch": 2.0637450199203187,
+      "grad_norm": 2.8208255767822266,
+      "learning_rate": 2.8278649599418194e-06,
+      "loss": 0.5798797607421875,
+      "step": 518
+    },
+    {
+      "epoch": 2.0717131474103585,
+      "grad_norm": 0.8215383291244507,
+      "learning_rate": 2.813037877720683e-06,
+      "loss": 0.9158266186714172,
+      "step": 520
+    },
+    {
+      "epoch": 2.0796812749003983,
+      "grad_norm": 0.9719216823577881,
+      "learning_rate": 2.798208055352017e-06,
+      "loss": 1.0235378742218018,
+      "step": 522
+    },
+    {
+      "epoch": 2.087649402390438,
+      "grad_norm": 0.9294026494026184,
+      "learning_rate": 2.7833761374625546e-06,
+      "loss": 1.0457916259765625,
+      "step": 524
+    },
+    {
+      "epoch": 2.095617529880478,
+      "grad_norm": 0.5908783078193665,
+      "learning_rate": 2.7685427687701172e-06,
+      "loss": 0.41702112555503845,
+      "step": 526
+    },
+    {
+      "epoch": 2.103585657370518,
+      "grad_norm": 0.5514804720878601,
+      "learning_rate": 2.753708594055588e-06,
+      "loss": 0.7837967872619629,
+      "step": 528
+    },
+    {
+      "epoch": 2.1115537848605577,
+      "grad_norm": 0.2083100825548172,
+      "learning_rate": 2.7388742581348893e-06,
+      "loss": 0.7011261582374573,
+      "step": 530
+    },
+    {
+      "epoch": 2.1195219123505975,
+      "grad_norm": 0.2741304934024811,
+      "learning_rate": 2.7240404058309487e-06,
+      "loss": 0.2263403981924057,
+      "step": 532
+    },
+    {
+      "epoch": 2.1274900398406373,
+      "grad_norm": 6.789957523345947,
+      "learning_rate": 2.709207681945673e-06,
+      "loss": 0.8779645562171936,
+      "step": 534
+    },
+    {
+      "epoch": 2.135458167330677,
+      "grad_norm": 1.5695756673812866,
+      "learning_rate": 2.6943767312319167e-06,
+      "loss": 0.693030595779419,
+      "step": 536
+    },
+    {
+      "epoch": 2.143426294820717,
+      "grad_norm": 5.31996488571167,
+      "learning_rate": 2.679548198365462e-06,
+      "loss": 0.670174777507782,
+      "step": 538
+    },
+    {
+      "epoch": 2.151394422310757,
+      "grad_norm": 0.5753666758537292,
+      "learning_rate": 2.6647227279169873e-06,
+      "loss": 0.938621997833252,
+      "step": 540
+    },
+    {
+      "epoch": 2.1593625498007967,
+      "grad_norm": 0.4965355694293976,
+      "learning_rate": 2.6499009643240527e-06,
+      "loss": 0.9415785670280457,
+      "step": 542
+    },
+    {
+      "epoch": 2.1673306772908365,
+      "grad_norm": 0.5600335597991943,
+      "learning_rate": 2.6350835518630907e-06,
+      "loss": 0.7928354144096375,
+      "step": 544
+    },
+    {
+      "epoch": 2.1752988047808763,
+      "grad_norm": 0.6061099767684937,
+      "learning_rate": 2.620271134621397e-06,
+      "loss": 0.7019538283348083,
+      "step": 546
+    },
+    {
+      "epoch": 2.183266932270916,
+      "grad_norm": 0.5260950922966003,
+      "learning_rate": 2.605464356469131e-06,
+      "loss": 0.9919465184211731,
+      "step": 548
+    },
+    {
+      "epoch": 2.191235059760956,
+      "grad_norm": 0.7529751658439636,
+      "learning_rate": 2.5906638610313336e-06,
+      "loss": 1.0579588413238525,
+      "step": 550
+    },
+    {
+      "epoch": 2.199203187250996,
+      "grad_norm": 0.5573561191558838,
+      "learning_rate": 2.575870291659946e-06,
+      "loss": 0.7870108485221863,
+      "step": 552
+    },
+    {
+      "epoch": 2.2071713147410357,
+      "grad_norm": 1.534278392791748,
+      "learning_rate": 2.5610842914058446e-06,
+      "loss": 0.4587668478488922,
+      "step": 554
+    },
+    {
+      "epoch": 2.2151394422310755,
+      "grad_norm": 0.5446329712867737,
+      "learning_rate": 2.5463065029908906e-06,
+      "loss": 0.6884219646453857,
+      "step": 556
+    },
+    {
+      "epoch": 2.2231075697211153,
+      "grad_norm": 0.950456976890564,
+      "learning_rate": 2.5315375687799896e-06,
+      "loss": 0.8508890867233276,
+      "step": 558
+    },
+    {
+      "epoch": 2.231075697211155,
+      "grad_norm": 1.1984989643096924,
+      "learning_rate": 2.5167781307531714e-06,
+      "loss": 0.9456567168235779,
+      "step": 560
+    },
+    {
+      "epoch": 2.239043824701195,
+      "grad_norm": 0.5195342302322388,
+      "learning_rate": 2.5020288304776822e-06,
+      "loss": 1.0414419174194336,
+      "step": 562
+    },
+    {
+      "epoch": 2.247011952191235,
+      "grad_norm": 1.8298522233963013,
+      "learning_rate": 2.487290309080097e-06,
+      "loss": 0.8636091351509094,
+      "step": 564
+    },
+    {
+      "epoch": 2.2549800796812747,
+      "grad_norm": 0.4848536252975464,
+      "learning_rate": 2.472563207218454e-06,
+      "loss": 0.7028620839118958,
+      "step": 566
+    },
+    {
+      "epoch": 2.2629482071713145,
+      "grad_norm": 0.5324665307998657,
+      "learning_rate": 2.457848165054402e-06,
+      "loss": 0.746355414390564,
+      "step": 568
+    },
+    {
+      "epoch": 2.2709163346613543,
+      "grad_norm": 0.32663580775260925,
+      "learning_rate": 2.4431458222253755e-06,
+      "loss": 1.0267982482910156,
+      "step": 570
+    },
+    {
+      "epoch": 2.278884462151394,
+      "grad_norm": 1.593503475189209,
+      "learning_rate": 2.4284568178167923e-06,
+      "loss": 0.6111531853675842,
+      "step": 572
+    },
+    {
+      "epoch": 2.2868525896414345,
+      "grad_norm": 0.40522316098213196,
+      "learning_rate": 2.413781790334272e-06,
+      "loss": 0.7418013215065002,
+      "step": 574
+    },
+    {
+      "epoch": 2.2948207171314743,
+      "grad_norm": 0.8094971776008606,
+      "learning_rate": 2.3991213776758776e-06,
+      "loss": 0.24140417575836182,
+      "step": 576
+    },
+    {
+      "epoch": 2.302788844621514,
+      "grad_norm": 0.732704758644104,
+      "learning_rate": 2.3844762171043975e-06,
+      "loss": 0.644301176071167,
+      "step": 578
+    },
+    {
+      "epoch": 2.310756972111554,
+      "grad_norm": 0.5985328555107117,
+      "learning_rate": 2.3698469452196328e-06,
+      "loss": 0.6490839123725891,
+      "step": 580
+    },
+    {
+      "epoch": 2.318725099601594,
+      "grad_norm": 1.812828779220581,
+      "learning_rate": 2.355234197930735e-06,
+      "loss": 0.6858562231063843,
+      "step": 582
+    },
+    {
+      "epoch": 2.3266932270916336,
+      "grad_norm": 0.35621851682662964,
+      "learning_rate": 2.340638610428554e-06,
+      "loss": 0.7984070777893066,
+      "step": 584
+    },
+    {
+      "epoch": 2.3346613545816735,
+      "grad_norm": 0.7255008816719055,
+      "learning_rate": 2.3260608171580412e-06,
+      "loss": 0.6289466023445129,
+      "step": 586
+    },
+    {
+      "epoch": 2.3426294820717133,
+      "grad_norm": 2.085850238800049,
+      "learning_rate": 2.311501451790658e-06,
+      "loss": 0.8911991715431213,
+      "step": 588
+    },
+    {
+      "epoch": 2.350597609561753,
+      "grad_norm": 0.5989985466003418,
+      "learning_rate": 2.2969611471968386e-06,
+      "loss": 1.017614722251892,
+      "step": 590
+    },
+    {
+      "epoch": 2.358565737051793,
+      "grad_norm": 0.9773586988449097,
+      "learning_rate": 2.2824405354184776e-06,
+      "loss": 1.049248456954956,
+      "step": 592
+    },
+    {
+      "epoch": 2.366533864541833,
+      "grad_norm": 0.5815195441246033,
+      "learning_rate": 2.267940247641458e-06,
+      "loss": 0.7822281718254089,
+      "step": 594
+    },
+    {
+      "epoch": 2.3745019920318726,
+      "grad_norm": 0.4693529009819031,
+      "learning_rate": 2.2534609141682135e-06,
+      "loss": 0.9958940744400024,
+      "step": 596
+    },
+    {
+      "epoch": 2.3824701195219125,
+      "grad_norm": 0.9353572130203247,
+      "learning_rate": 2.2390031643903286e-06,
+      "loss": 0.7725387811660767,
+      "step": 598
+    },
+    {
+      "epoch": 2.3904382470119523,
+      "grad_norm": 8.899816513061523,
+      "learning_rate": 2.224567626761185e-06,
+      "loss": 0.9183634519577026,
+      "step": 600
+    },
+    {
+      "epoch": 2.398406374501992,
+      "grad_norm": 1.1901518106460571,
+      "learning_rate": 2.2101549287686384e-06,
+      "loss": 1.0622332096099854,
+      "step": 602
+    },
+    {
+      "epoch": 2.406374501992032,
+      "grad_norm": 0.2631213068962097,
+      "learning_rate": 2.195765696907745e-06,
+      "loss": 0.5778547525405884,
+      "step": 604
+    },
+    {
+      "epoch": 2.414342629482072,
+      "grad_norm": 1.9375444650650024,
+      "learning_rate": 2.1814005566535303e-06,
+      "loss": 0.27354422211647034,
+      "step": 606
+    },
+    {
+      "epoch": 2.4223107569721116,
+      "grad_norm": 0.818555474281311,
+      "learning_rate": 2.167060132433798e-06,
+      "loss": 0.591134786605835,
+      "step": 608
+    },
+    {
+      "epoch": 2.4302788844621515,
+      "grad_norm": 0.5963910818099976,
+      "learning_rate": 2.1527450476019895e-06,
+      "loss": 0.9486387968063354,
+      "step": 610
+    },
+    {
+      "epoch": 2.4382470119521913,
+      "grad_norm": 0.39765867590904236,
+      "learning_rate": 2.138455924410085e-06,
+      "loss": 0.7022622227668762,
+      "step": 612
+    },
+    {
+      "epoch": 2.446215139442231,
+      "grad_norm": 0.5631069540977478,
+      "learning_rate": 2.1241933839815606e-06,
+      "loss": 0.5781379342079163,
+      "step": 614
+    },
+    {
+      "epoch": 2.454183266932271,
+      "grad_norm": 0.49549999833106995,
+      "learning_rate": 2.109958046284382e-06,
+      "loss": 0.9160446524620056,
+      "step": 616
+    },
+    {
+      "epoch": 2.462151394422311,
+      "grad_norm": 0.6964736580848694,
+      "learning_rate": 2.0957505301040604e-06,
+      "loss": 0.7290208339691162,
+      "step": 618
+    },
+    {
+      "epoch": 2.4701195219123506,
+      "grad_norm": 3.0886189937591553,
+      "learning_rate": 2.081571453016756e-06,
+      "loss": 0.8496452569961548,
+      "step": 620
+    },
+    {
+      "epoch": 2.4780876494023905,
+      "grad_norm": 0.5862022638320923,
+      "learning_rate": 2.067421431362427e-06,
+      "loss": 0.8492017984390259,
+      "step": 622
+    },
+    {
+      "epoch": 2.4860557768924303,
+      "grad_norm": 0.9748086929321289,
+      "learning_rate": 2.053301080218044e-06,
+      "loss": 0.7983825206756592,
+      "step": 624
+    },
+    {
+      "epoch": 2.49402390438247,
+      "grad_norm": 0.9882903695106506,
+      "learning_rate": 2.039211013370853e-06,
+      "loss": 0.6598578095436096,
+      "step": 626
+    },
+    {
+      "epoch": 2.50199203187251,
+      "grad_norm": 0.6953760385513306,
+      "learning_rate": 2.025151843291693e-06,
+      "loss": 1.0048404932022095,
+      "step": 628
+    },
+    {
+      "epoch": 2.50996015936255,
+      "grad_norm": 0.7751678824424744,
+      "learning_rate": 2.0111241811083733e-06,
+      "loss": 0.5011482834815979,
+      "step": 630
+    },
+    {
+      "epoch": 2.5179282868525896,
+      "grad_norm": 1.8852907419204712,
+      "learning_rate": 1.9971286365791092e-06,
+      "loss": 0.47957643866539,
+      "step": 632
+    },
+    {
+      "epoch": 2.5258964143426295,
+      "grad_norm": 0.8766360878944397,
+      "learning_rate": 1.9831658180660197e-06,
+      "loss": 1.0223110914230347,
+      "step": 634
+    },
+    {
+      "epoch": 2.5338645418326693,
+      "grad_norm": 0.4993748068809509,
+      "learning_rate": 1.9692363325086774e-06,
+      "loss": 0.9816610813140869,
+      "step": 636
+    },
+    {
+      "epoch": 2.541832669322709,
+      "grad_norm": 3.096184253692627,
+      "learning_rate": 1.95534078539773e-06,
+      "loss": 0.994441032409668,
+      "step": 638
+    },
+    {
+      "epoch": 2.549800796812749,
+      "grad_norm": 0.25416290760040283,
+      "learning_rate": 1.9414797807485796e-06,
+      "loss": 0.7792674899101257,
+      "step": 640
+    },
+    {
+      "epoch": 2.557768924302789,
+      "grad_norm": 0.5611869096755981,
+      "learning_rate": 1.9276539210751295e-06,
+      "loss": 0.9999116659164429,
+      "step": 642
+    },
+    {
+      "epoch": 2.5657370517928286,
+      "grad_norm": 0.4883551299571991,
+      "learning_rate": 1.9138638073635886e-06,
+      "loss": 1.0858542919158936,
+      "step": 644
+    },
+    {
+      "epoch": 2.5737051792828685,
+      "grad_norm": 0.5001575350761414,
+      "learning_rate": 1.90011003904635e-06,
+      "loss": 0.5845852494239807,
+      "step": 646
+    },
+    {
+      "epoch": 2.5816733067729083,
+      "grad_norm": 0.9940570592880249,
+      "learning_rate": 1.886393213975939e-06,
+      "loss": 1.0281263589859009,
+      "step": 648
+    },
+    {
+      "epoch": 2.589641434262948,
+      "grad_norm": 0.4883127808570862,
+      "learning_rate": 1.8727139283990203e-06,
+      "loss": 0.7996731996536255,
+      "step": 650
+    },
+    {
+      "epoch": 2.597609561752988,
+      "grad_norm": 0.6091388463973999,
+      "learning_rate": 1.8590727769304789e-06,
+      "loss": 0.849186897277832,
+      "step": 652
+    },
+    {
+      "epoch": 2.605577689243028,
+      "grad_norm": 0.5755348205566406,
+      "learning_rate": 1.8454703525275814e-06,
+      "loss": 0.689119815826416,
+      "step": 654
+    },
+    {
+      "epoch": 2.6135458167330676,
+      "grad_norm": 0.7240902185440063,
+      "learning_rate": 1.8319072464641912e-06,
+      "loss": 0.35995274782180786,
+      "step": 656
+    },
+    {
+      "epoch": 2.6215139442231075,
+      "grad_norm": 0.6905495524406433,
+      "learning_rate": 1.8183840483050741e-06,
+      "loss": 0.5684748888015747,
+      "step": 658
+    },
+    {
+      "epoch": 2.6294820717131473,
+      "grad_norm": 0.9130073189735413,
+      "learning_rate": 1.8049013458802667e-06,
+      "loss": 0.8855925798416138,
+      "step": 660
+    },
+    {
+      "epoch": 2.637450199203187,
+      "grad_norm": 1.0199811458587646,
+      "learning_rate": 1.7914597252595267e-06,
+      "loss": 1.000173807144165,
+      "step": 662
+    },
+    {
+      "epoch": 2.645418326693227,
+      "grad_norm": 0.795620858669281,
+      "learning_rate": 1.7780597707268582e-06,
+      "loss": 0.579514741897583,
+      "step": 664
+    },
+    {
+      "epoch": 2.653386454183267,
+      "grad_norm": 0.783987283706665,
+      "learning_rate": 1.7647020647551105e-06,
+      "loss": 0.9255267977714539,
+      "step": 666
+    },
+    {
+      "epoch": 2.6613545816733066,
+      "grad_norm": 0.17200644314289093,
+      "learning_rate": 1.7513871879806605e-06,
+      "loss": 0.5811091661453247,
+      "step": 668
+    },
+    {
+      "epoch": 2.6693227091633465,
+      "grad_norm": 0.6199663281440735,
+      "learning_rate": 1.7381157191781766e-06,
+      "loss": 0.7342613339424133,
+      "step": 670
+    },
+    {
+      "epoch": 2.6772908366533863,
+      "grad_norm": 1.260032296180725,
+      "learning_rate": 1.7248882352354571e-06,
+      "loss": 0.37697556614875793,
+      "step": 672
+    },
+    {
+      "epoch": 2.685258964143426,
+      "grad_norm": 0.3898491859436035,
+      "learning_rate": 1.7117053111283528e-06,
+      "loss": 1.0542317628860474,
+      "step": 674
+    },
+    {
+      "epoch": 2.6932270916334664,
+      "grad_norm": 1.2614471912384033,
+      "learning_rate": 1.6985675198957752e-06,
+      "loss": 0.7127259373664856,
+      "step": 676
+    },
+    {
+      "epoch": 2.7011952191235062,
+      "grad_norm": 0.6190806031227112,
+      "learning_rate": 1.6854754326147916e-06,
+      "loss": 0.6188024282455444,
+      "step": 678
+    },
+    {
+      "epoch": 2.709163346613546,
+      "grad_norm": 0.37103477120399475,
+      "learning_rate": 1.6724296183757925e-06,
+      "loss": 1.0318716764450073,
+      "step": 680
+    },
+    {
+      "epoch": 2.717131474103586,
+      "grad_norm": 6.4781012535095215,
+      "learning_rate": 1.6594306442577629e-06,
+      "loss": 1.0529109239578247,
+      "step": 682
+    },
+    {
+      "epoch": 2.7250996015936257,
+      "grad_norm": 0.8956589698791504,
+      "learning_rate": 1.6464790753036298e-06,
+      "loss": 0.6906982660293579,
+      "step": 684
+    },
+    {
+      "epoch": 2.7330677290836656,
+      "grad_norm": 2.446382761001587,
+      "learning_rate": 1.6335754744956935e-06,
+      "loss": 0.48746657371520996,
+      "step": 686
+    },
+    {
+      "epoch": 2.7410358565737054,
+      "grad_norm": 0.434888631105423,
+      "learning_rate": 1.6207204027311685e-06,
+      "loss": 1.0371174812316895,
+      "step": 688
+    },
+    {
+      "epoch": 2.7490039840637452,
+      "grad_norm": 1.4846906661987305,
+      "learning_rate": 1.6079144187977925e-06,
+      "loss": 0.5729243159294128,
+      "step": 690
+    },
+    {
+      "epoch": 2.756972111553785,
+      "grad_norm": 0.8365349769592285,
+      "learning_rate": 1.595158079349542e-06,
+      "loss": 1.1043401956558228,
+      "step": 692
+    },
+    {
+      "epoch": 2.764940239043825,
+      "grad_norm": 1.0761866569519043,
+      "learning_rate": 1.5824519388824311e-06,
+      "loss": 0.8631709218025208,
+      "step": 694
+    },
+    {
+      "epoch": 2.7729083665338647,
+      "grad_norm": 0.9276056289672852,
+      "learning_rate": 1.5697965497104117e-06,
+      "loss": 0.7128748893737793,
+      "step": 696
+    },
+    {
+      "epoch": 2.7808764940239046,
+      "grad_norm": 0.6187699437141418,
+      "learning_rate": 1.557192461941367e-06,
+      "loss": 0.8992443680763245,
+      "step": 698
+    },
+    {
+      "epoch": 2.7888446215139444,
+      "grad_norm": 0.39032015204429626,
+      "learning_rate": 1.5446402234531934e-06,
+      "loss": 1.0499823093414307,
+      "step": 700
+    },
+    {
+      "epoch": 2.7968127490039842,
+      "grad_norm": 0.5455421805381775,
+      "learning_rate": 1.5321403798699924e-06,
+      "loss": 0.6968839168548584,
+      "step": 702
+    },
+    {
+      "epoch": 2.804780876494024,
+      "grad_norm": 1.6619435548782349,
+      "learning_rate": 1.5196934745383456e-06,
+      "loss": 1.226813793182373,
+      "step": 704
+    },
+    {
+      "epoch": 2.812749003984064,
+      "grad_norm": 0.42876794934272766,
+      "learning_rate": 1.5073000485037044e-06,
+      "loss": 1.2229830026626587,
+      "step": 706
+    },
+    {
+      "epoch": 2.8207171314741037,
+      "grad_norm": 1.3915983438491821,
+      "learning_rate": 1.4949606404868638e-06,
+      "loss": 1.093475103378296,
+      "step": 708
+    },
+    {
+      "epoch": 2.8286852589641436,
+      "grad_norm": 0.8409278392791748,
+      "learning_rate": 1.4826757868605519e-06,
+      "loss": 0.5333719253540039,
+      "step": 710
+    },
+    {
+      "epoch": 2.8366533864541834,
+      "grad_norm": 0.7415921092033386,
+      "learning_rate": 1.4704460216261125e-06,
+      "loss": 1.0662540197372437,
+      "step": 712
+    },
+    {
+      "epoch": 2.8446215139442232,
+      "grad_norm": 0.5356712937355042,
+      "learning_rate": 1.458271876390287e-06,
+      "loss": 0.6107707023620605,
+      "step": 714
+    },
+    {
+      "epoch": 2.852589641434263,
+      "grad_norm": 0.4340842664241791,
+      "learning_rate": 1.446153880342117e-06,
+      "loss": 1.031813144683838,
+      "step": 716
+    },
+    {
+      "epoch": 2.860557768924303,
+      "grad_norm": 0.7240418791770935,
+      "learning_rate": 1.4340925602299337e-06,
+      "loss": 0.6472567319869995,
+      "step": 718
+    },
+    {
+      "epoch": 2.8685258964143427,
+      "grad_norm": 0.5913697481155396,
+      "learning_rate": 1.4220884403384656e-06,
+      "loss": 0.5840514898300171,
+      "step": 720
+    },
+    {
+      "epoch": 2.8764940239043826,
+      "grad_norm": 0.9075729250907898,
+      "learning_rate": 1.4101420424660433e-06,
+      "loss": 0.6979702115058899,
+      "step": 722
+    },
+    {
+      "epoch": 2.8844621513944224,
+      "grad_norm": 0.34114325046539307,
+      "learning_rate": 1.3982538859019219e-06,
+      "loss": 0.6354752779006958,
+      "step": 724
+    },
+    {
+      "epoch": 2.8924302788844622,
+      "grad_norm": 0.612322211265564,
+      "learning_rate": 1.3864244874037096e-06,
+      "loss": 0.992059588432312,
+      "step": 726
+    },
+    {
+      "epoch": 2.900398406374502,
+      "grad_norm": 0.9481393694877625,
+      "learning_rate": 1.3746543611749007e-06,
+      "loss": 0.5739765167236328,
+      "step": 728
+    },
+    {
+      "epoch": 2.908366533864542,
+      "grad_norm": 0.6491210460662842,
+      "learning_rate": 1.3629440188425285e-06,
+      "loss": 0.9585099816322327,
+      "step": 730
+    },
+    {
+      "epoch": 2.9163346613545817,
+      "grad_norm": 2.533921003341675,
+      "learning_rate": 1.351293969434921e-06,
+      "loss": 0.40510398149490356,
+      "step": 732
+    },
+    {
+      "epoch": 2.9243027888446216,
+      "grad_norm": 0.4380839169025421,
+      "learning_rate": 1.3397047193595809e-06,
+      "loss": 1.0545580387115479,
+      "step": 734
+    },
+    {
+      "epoch": 2.9322709163346614,
+      "grad_norm": 1.7832928895950317,
+      "learning_rate": 1.3281767723811651e-06,
+      "loss": 0.46457114815711975,
+      "step": 736
+    },
+    {
+      "epoch": 2.9402390438247012,
+      "grad_norm": 3.623868942260742,
+      "learning_rate": 1.3167106295995923e-06,
+      "loss": 0.9381816387176514,
+      "step": 738
+    },
+    {
+      "epoch": 2.948207171314741,
+      "grad_norm": 0.6202852129936218,
+      "learning_rate": 1.3053067894282624e-06,
+      "loss": 1.1592164039611816,
+      "step": 740
+    },
+    {
+      "epoch": 2.956175298804781,
+      "grad_norm": 1.329702377319336,
+      "learning_rate": 1.2939657475723844e-06,
+      "loss": 0.445529043674469,
+      "step": 742
+    },
+    {
+      "epoch": 2.9641434262948207,
+      "grad_norm": 0.41946062445640564,
+      "learning_rate": 1.2826879970074346e-06,
+      "loss": 0.9050129652023315,
+      "step": 744
+    },
+    {
+      "epoch": 2.9721115537848606,
+      "grad_norm": 0.5631400346755981,
+      "learning_rate": 1.271474027957727e-06,
+      "loss": 0.8985243439674377,
+      "step": 746
+    },
+    {
+      "epoch": 2.9800796812749004,
+      "grad_norm": 0.9848307371139526,
+      "learning_rate": 1.2603243278751043e-06,
+      "loss": 0.5436593890190125,
+      "step": 748
+    },
+    {
+      "epoch": 2.9880478087649402,
+      "grad_norm": 0.4108133614063263,
+      "learning_rate": 1.249239381417746e-06,
+      "loss": 0.14398813247680664,
+      "step": 750
+    },
+    {
+      "epoch": 2.99601593625498,
+      "grad_norm": 1.5209935903549194,
+      "learning_rate": 1.238219670429104e-06,
+      "loss": 1.0190974473953247,
+      "step": 752
+    },
+    {
+      "epoch": 3.00398406374502,
+      "grad_norm": 0.6851494908332825,
+      "learning_rate": 1.2272656739169589e-06,
+      "loss": 0.551688551902771,
+      "step": 754
+    },
+    {
+      "epoch": 3.0119521912350598,
+      "grad_norm": 0.6992303133010864,
+      "learning_rate": 1.2163778680325948e-06,
+      "loss": 0.6257609128952026,
+      "step": 756
+    },
+    {
+      "epoch": 3.0199203187250996,
+      "grad_norm": 0.42870357632637024,
+      "learning_rate": 1.2055567260501055e-06,
+      "loss": 0.3591914474964142,
+      "step": 758
+    },
+    {
+      "epoch": 3.0278884462151394,
+      "grad_norm": 0.6524999141693115,
+      "learning_rate": 1.1948027183458182e-06,
+      "loss": 0.6760011911392212,
+      "step": 760
+    },
+    {
+      "epoch": 3.0358565737051793,
+      "grad_norm": 1.3119370937347412,
+      "learning_rate": 1.18411631237785e-06,
+      "loss": 0.5357263684272766,
+      "step": 762
+    },
+    {
+      "epoch": 3.043824701195219,
+      "grad_norm": 0.6288226246833801,
+      "learning_rate": 1.173497972665788e-06,
+      "loss": 0.9737890958786011,
+      "step": 764
+    },
+    {
+      "epoch": 3.051792828685259,
+      "grad_norm": 1.2268046140670776,
+      "learning_rate": 1.1629481607704964e-06,
+      "loss": 0.16091720759868622,
+      "step": 766
+    },
+    {
+      "epoch": 3.0597609561752988,
+      "grad_norm": 0.5511128902435303,
+      "learning_rate": 1.1524673352740543e-06,
+      "loss": 0.5842117071151733,
+      "step": 768
+    },
+    {
+      "epoch": 3.0677290836653386,
+      "grad_norm": 0.5404235124588013,
+      "learning_rate": 1.1420559517598206e-06,
+      "loss": 0.806337296962738,
+      "step": 770
+    },
+    {
+      "epoch": 3.0756972111553784,
+      "grad_norm": 0.6832277774810791,
+      "learning_rate": 1.13171446279263e-06,
+      "loss": 0.9935510158538818,
+      "step": 772
+    },
+    {
+      "epoch": 3.0836653386454183,
+      "grad_norm": 2.003978729248047,
+      "learning_rate": 1.1214433178991258e-06,
+      "loss": 0.909023642539978,
+      "step": 774
+    },
+    {
+      "epoch": 3.091633466135458,
+      "grad_norm": 0.2051975131034851,
+      "learning_rate": 1.111242963548212e-06,
+      "loss": 0.5670095682144165,
+      "step": 776
+    },
+    {
+      "epoch": 3.099601593625498,
+      "grad_norm": 0.6747473478317261,
+      "learning_rate": 1.1011138431316538e-06,
+      "loss": 0.5940138697624207,
+      "step": 778
+    },
+    {
+      "epoch": 3.1075697211155378,
+      "grad_norm": 0.4947768747806549,
+      "learning_rate": 1.0910563969447969e-06,
+      "loss": 1.0205563306808472,
+      "step": 780
+    },
+    {
+      "epoch": 3.1155378486055776,
+      "grad_norm": 0.175948828458786,
+      "learning_rate": 1.0810710621674355e-06,
+      "loss": 0.3414739668369293,
+      "step": 782
+    },
+    {
+      "epoch": 3.1235059760956174,
+      "grad_norm": 0.5083895921707153,
+      "learning_rate": 1.0711582728448034e-06,
+      "loss": 1.019008994102478,
+      "step": 784
+    },
+    {
+      "epoch": 3.1314741035856573,
+      "grad_norm": 0.41630658507347107,
+      "learning_rate": 1.0613184598687088e-06,
+      "loss": 0.8115735054016113,
+      "step": 786
+    },
+    {
+      "epoch": 3.139442231075697,
+      "grad_norm": 35.31654357910156,
+      "learning_rate": 1.0515520509588076e-06,
+      "loss": 0.6538347005844116,
+      "step": 788
+    },
+    {
+      "epoch": 3.147410358565737,
+      "grad_norm": 0.5832368731498718,
+      "learning_rate": 1.0418594706440022e-06,
+      "loss": 0.545616626739502,
+      "step": 790
+    },
+    {
+      "epoch": 3.1553784860557768,
+      "grad_norm": 1.0365172624588013,
+      "learning_rate": 1.0322411402439982e-06,
+      "loss": 0.7904989123344421,
+      "step": 792
+    },
+    {
+      "epoch": 3.1633466135458166,
+      "grad_norm": 0.10791326314210892,
+      "learning_rate": 1.0226974778509846e-06,
+      "loss": 0.279825359582901,
+      "step": 794
+    },
+    {
+      "epoch": 3.1713147410358564,
+      "grad_norm": 0.911193311214447,
+      "learning_rate": 1.0132288983114611e-06,
+      "loss": 0.6512662172317505,
+      "step": 796
+    },
+    {
+      "epoch": 3.1792828685258963,
+      "grad_norm": 1.0532306432724,
+      "learning_rate": 1.003835813208206e-06,
+      "loss": 0.553152322769165,
+      "step": 798
+    },
+    {
+      "epoch": 3.187250996015936,
+      "grad_norm": 0.5105674862861633,
+      "learning_rate": 9.945186308423832e-07,
+      "loss": 0.8661262392997742,
+      "step": 800
+    },
+    {
+      "epoch": 3.195219123505976,
+      "grad_norm": 0.4801194369792938,
+      "learning_rate": 9.852777562157983e-07,
+      "loss": 0.7538309097290039,
+      "step": 802
+    },
+    {
+      "epoch": 3.2031872509960158,
+      "grad_norm": 1.5633740425109863,
+      "learning_rate": 9.7611359101329e-07,
+      "loss": 0.6893703937530518,
+      "step": 804
+    },
+    {
+      "epoch": 3.2111553784860556,
+      "grad_norm": 1.099211573600769,
+      "learning_rate": 9.670265335852725e-07,
+      "loss": 1.1113306283950806,
+      "step": 806
+    },
+    {
+      "epoch": 3.2191235059760954,
+      "grad_norm": 1.2671184539794922,
+      "learning_rate": 9.580169789304152e-07,
+      "loss": 0.4243836998939514,
+      "step": 808
+    },
+    {
+      "epoch": 3.2270916334661353,
+      "grad_norm": 1.1165072917938232,
+      "learning_rate": 9.490853186784801e-07,
+      "loss": 0.7039945721626282,
+      "step": 810
+    },
+    {
+      "epoch": 3.235059760956175,
+      "grad_norm": 1.370728850364685,
+      "learning_rate": 9.402319410732911e-07,
+      "loss": 0.3937326669692993,
+      "step": 812
+    },
+    {
+      "epoch": 3.243027888446215,
+      "grad_norm": 0.5677525997161865,
+      "learning_rate": 9.314572309558617e-07,
+      "loss": 0.8592427372932434,
+      "step": 814
+    },
+    {
+      "epoch": 3.2509960159362548,
+      "grad_norm": 2.0334928035736084,
+      "learning_rate": 9.227615697476674e-07,
+      "loss": 0.32591021060943604,
+      "step": 816
+    },
+    {
+      "epoch": 3.2589641434262946,
+      "grad_norm": 0.5305483341217041,
+      "learning_rate": 9.141453354340619e-07,
+      "loss": 0.9562654495239258,
+      "step": 818
+    },
+    {
+      "epoch": 3.2669322709163344,
+      "grad_norm": 2.4839959144592285,
+      "learning_rate": 9.056089025478495e-07,
+      "loss": 0.4911564290523529,
+      "step": 820
+    },
+    {
+      "epoch": 3.2749003984063743,
+      "grad_norm": 2.0328733921051025,
+      "learning_rate": 8.971526421530067e-07,
+      "loss": 0.8154242038726807,
+      "step": 822
+    },
+    {
+      "epoch": 3.2828685258964145,
+      "grad_norm": 0.7321983575820923,
+      "learning_rate": 8.887769218285469e-07,
+      "loss": 0.5709739923477173,
+      "step": 824
+    },
+    {
+      "epoch": 3.2908366533864544,
+      "grad_norm": 2.088676929473877,
+      "learning_rate": 8.804821056525498e-07,
+      "loss": 0.971275269985199,
+      "step": 826
+    },
+    {
+      "epoch": 3.298804780876494,
+      "grad_norm": 0.7342613339424133,
+      "learning_rate": 8.722685541863284e-07,
+      "loss": 0.9591888189315796,
+      "step": 828
+    },
+    {
+      "epoch": 3.306772908366534,
+      "grad_norm": 0.9373747706413269,
+      "learning_rate": 8.641366244587612e-07,
+      "loss": 0.9429539442062378,
+      "step": 830
+    },
+    {
+      "epoch": 3.314741035856574,
+      "grad_norm": 3.6311287879943848,
+      "learning_rate": 8.560866699507703e-07,
+      "loss": 0.47473788261413574,
+      "step": 832
+    },
+    {
+      "epoch": 3.3227091633466137,
+      "grad_norm": 1.8518421649932861,
+      "learning_rate": 8.481190405799569e-07,
+      "loss": 0.3490409851074219,
+      "step": 834
+    },
+    {
+      "epoch": 3.3306772908366535,
+      "grad_norm": 1.1139072179794312,
+      "learning_rate": 8.402340826853908e-07,
+      "loss": 0.551905632019043,
+      "step": 836
+    },
+    {
+      "epoch": 3.3386454183266934,
+      "grad_norm": 1.1520075798034668,
+      "learning_rate": 8.324321390125548e-07,
+      "loss": 0.8061926960945129,
+      "step": 838
+    },
+    {
+      "epoch": 3.346613545816733,
+      "grad_norm": 1.0629056692123413,
+      "learning_rate": 8.247135486984487e-07,
+      "loss": 0.20331330597400665,
+      "step": 840
+    },
+    {
+      "epoch": 3.354581673306773,
+      "grad_norm": 0.7650991082191467,
+      "learning_rate": 8.170786472568451e-07,
+      "loss": 0.7244747281074524,
+      "step": 842
+    },
+    {
+      "epoch": 3.362549800796813,
+      "grad_norm": 0.7189844846725464,
+      "learning_rate": 8.095277665637061e-07,
+      "loss": 0.6685651540756226,
+      "step": 844
+    },
+    {
+      "epoch": 3.3705179282868527,
+      "grad_norm": 0.4765610694885254,
+      "learning_rate": 8.020612348427572e-07,
+      "loss": 1.0068601369857788,
+      "step": 846
+    },
+    {
+      "epoch": 3.3784860557768925,
+      "grad_norm": 0.540550708770752,
+      "learning_rate": 7.946793766512195e-07,
+      "loss": 0.9561856985092163,
+      "step": 848
+    },
+    {
+      "epoch": 3.3864541832669324,
+      "grad_norm": 1.3491733074188232,
+      "learning_rate": 7.873825128657034e-07,
+      "loss": 1.0059118270874023,
+      "step": 850
+    },
+    {
+      "epoch": 3.394422310756972,
+      "grad_norm": 0.6583026051521301,
+      "learning_rate": 7.801709606682573e-07,
+      "loss": 0.9497774839401245,
+      "step": 852
+    },
+    {
+      "epoch": 3.402390438247012,
+      "grad_norm": 0.9283490777015686,
+      "learning_rate": 7.730450335325852e-07,
+      "loss": 0.577174723148346,
+      "step": 854
+    },
+    {
+      "epoch": 3.410358565737052,
+      "grad_norm": 0.814186155796051,
+      "learning_rate": 7.660050412104153e-07,
+      "loss": 0.9153207540512085,
+      "step": 856
+    },
+    {
+      "epoch": 3.4183266932270917,
+      "grad_norm": 3.568722724914551,
+      "learning_rate": 7.5905128971804e-07,
+      "loss": 1.0680537223815918,
+      "step": 858
+    },
+    {
+      "epoch": 3.4262948207171315,
+      "grad_norm": 3.574244737625122,
+      "learning_rate": 7.521840813230098e-07,
+      "loss": 1.0113861560821533,
+      "step": 860
+    },
+    {
+      "epoch": 3.4342629482071714,
+      "grad_norm": 0.4404396116733551,
+      "learning_rate": 7.454037145309986e-07,
+      "loss": 0.7681607604026794,
+      "step": 862
+    },
+    {
+      "epoch": 3.442231075697211,
+      "grad_norm": 1.1538301706314087,
+      "learning_rate": 7.38710484072826e-07,
+      "loss": 0.6256715059280396,
+      "step": 864
+    },
+    {
+      "epoch": 3.450199203187251,
+      "grad_norm": 0.7298967242240906,
+      "learning_rate": 7.321046808916427e-07,
+      "loss": 0.8676828145980835,
+      "step": 866
+    },
+    {
+      "epoch": 3.458167330677291,
+      "grad_norm": 2.115961790084839,
+      "learning_rate": 7.255865921302902e-07,
+      "loss": 0.7033858299255371,
+      "step": 868
+    },
+    {
+      "epoch": 3.4661354581673307,
+      "grad_norm": 1.0571824312210083,
+      "learning_rate": 7.191565011188149e-07,
+      "loss": 0.5359318852424622,
+      "step": 870
+    },
+    {
+      "epoch": 3.4741035856573705,
+      "grad_norm": 0.1546042263507843,
+      "learning_rate": 7.128146873621533e-07,
+      "loss": 0.5908284187316895,
+      "step": 872
+    },
+    {
+      "epoch": 3.4820717131474104,
+      "grad_norm": 0.2078041136264801,
+      "learning_rate": 7.065614265279819e-07,
+      "loss": 0.6999475359916687,
+      "step": 874
+    },
+    {
+      "epoch": 3.49003984063745,
+      "grad_norm": 0.4830763638019562,
+      "learning_rate": 7.00396990434734e-07,
+      "loss": 0.5590264201164246,
+      "step": 876
+    },
+    {
+      "epoch": 3.49800796812749,
+      "grad_norm": 0.47940587997436523,
+      "learning_rate": 6.943216470397872e-07,
+      "loss": 0.5570573806762695,
+      "step": 878
+    },
+    {
+      "epoch": 3.50597609561753,
+      "grad_norm": 0.20077168941497803,
+      "learning_rate": 6.883356604278116e-07,
+      "loss": 0.41967588663101196,
+      "step": 880
+    },
+    {
+      "epoch": 3.5139442231075697,
+      "grad_norm": 0.4751814901828766,
+      "learning_rate": 6.824392907992938e-07,
+      "loss": 0.6674134135246277,
+      "step": 882
+    },
+    {
+      "epoch": 3.5219123505976095,
+      "grad_norm": 1.6535459756851196,
+      "learning_rate": 6.76632794459224e-07,
+      "loss": 0.27134981751441956,
+      "step": 884
+    },
+    {
+      "epoch": 3.5298804780876494,
+      "grad_norm": 0.6872122287750244,
+      "learning_rate": 6.709164238059568e-07,
+      "loss": 0.8428699970245361,
+      "step": 886
+    },
+    {
+      "epoch": 3.537848605577689,
+      "grad_norm": 0.5006639957427979,
+      "learning_rate": 6.652904273202385e-07,
+      "loss": 0.9281488656997681,
+      "step": 888
+    },
+    {
+      "epoch": 3.545816733067729,
+      "grad_norm": 0.6955797076225281,
+      "learning_rate": 6.597550495544071e-07,
+      "loss": 0.5072586536407471,
+      "step": 890
+    },
+    {
+      "epoch": 3.553784860557769,
+      "grad_norm": 0.9870802164077759,
+      "learning_rate": 6.54310531121762e-07,
+      "loss": 0.9480372071266174,
+      "step": 892
+    },
+    {
+      "epoch": 3.5617529880478087,
+      "grad_norm": 1.0111428499221802,
+      "learning_rate": 6.489571086861027e-07,
+      "loss": 1.066458821296692,
+      "step": 894
+    },
+    {
+      "epoch": 3.5697211155378485,
+      "grad_norm": 0.8398950099945068,
+      "learning_rate": 6.436950149514448e-07,
+      "loss": 0.2645135819911957,
+      "step": 896
+    },
+    {
+      "epoch": 3.5776892430278884,
+      "grad_norm": 0.7592117190361023,
+      "learning_rate": 6.385244786519036e-07,
+      "loss": 0.6446539163589478,
+      "step": 898
+    },
+    {
+      "epoch": 3.585657370517928,
+      "grad_norm": 0.5307031273841858,
+      "learning_rate": 6.334457245417491e-07,
+      "loss": 0.2668192982673645,
+      "step": 900
+    },
+    {
+      "epoch": 3.593625498007968,
+      "grad_norm": 0.7873986959457397,
+      "learning_rate": 6.284589733856403e-07,
+      "loss": 0.09220651537179947,
+      "step": 902
+    },
+    {
+      "epoch": 3.601593625498008,
+      "grad_norm": 0.8216732740402222,
+      "learning_rate": 6.235644419490253e-07,
+      "loss": 0.7840216755867004,
+      "step": 904
+    },
+    {
+      "epoch": 3.6095617529880477,
+      "grad_norm": 0.6413154006004333,
+      "learning_rate": 6.18762342988722e-07,
+      "loss": 0.6411182284355164,
+      "step": 906
+    },
+    {
+      "epoch": 3.6175298804780875,
+      "grad_norm": 0.9323461055755615,
+      "learning_rate": 6.140528852436666e-07,
+      "loss": 0.9973750710487366,
+      "step": 908
+    },
+    {
+      "epoch": 3.6254980079681274,
+      "grad_norm": 0.9435186386108398,
+      "learning_rate": 6.094362734258438e-07,
+      "loss": 0.9649828672409058,
+      "step": 910
+    },
+    {
+      "epoch": 3.633466135458167,
+      "grad_norm": 0.3178688883781433,
+      "learning_rate": 6.049127082113853e-07,
+      "loss": 0.06291352957487106,
+      "step": 912
+    },
+    {
+      "epoch": 3.641434262948207,
+      "grad_norm": 0.8880316615104675,
+      "learning_rate": 6.004823862318475e-07,
+      "loss": 0.8320421576499939,
+      "step": 914
+    },
+    {
+      "epoch": 3.649402390438247,
+      "grad_norm": 5.795741558074951,
+      "learning_rate": 5.961455000656656e-07,
+      "loss": 0.8510156869888306,
+      "step": 916
+    },
+    {
+      "epoch": 3.6573705179282867,
+      "grad_norm": 3.2219347953796387,
+      "learning_rate": 5.919022382297808e-07,
+      "loss": 0.20207783579826355,
+      "step": 918
+    },
+    {
+      "epoch": 3.6653386454183265,
+      "grad_norm": 0.5910276770591736,
+      "learning_rate": 5.877527851714479e-07,
+      "loss": 0.957602858543396,
+      "step": 920
+    },
+    {
+      "epoch": 3.6733067729083664,
+      "grad_norm": 0.6868635416030884,
+      "learning_rate": 5.836973212602144e-07,
+      "loss": 0.17233799397945404,
+      "step": 922
+    },
+    {
+      "epoch": 3.681274900398406,
+      "grad_norm": 1.3047155141830444,
+      "learning_rate": 5.797360227800832e-07,
+      "loss": 1.018256425857544,
+      "step": 924
+    },
+    {
+      "epoch": 3.6892430278884465,
+      "grad_norm": 1.0058708190917969,
+      "learning_rate": 5.758690619218493e-07,
+      "loss": 0.3557318150997162,
+      "step": 926
+    },
+    {
+      "epoch": 3.6972111553784863,
+      "grad_norm": 0.8192755579948425,
+      "learning_rate": 5.720966067756131e-07,
+      "loss": 0.5570724010467529,
+      "step": 928
+    },
+    {
+      "epoch": 3.705179282868526,
+      "grad_norm": 0.3615506887435913,
+      "learning_rate": 5.684188213234763e-07,
+      "loss": 0.8607496023178101,
+      "step": 930
+    },
+    {
+      "epoch": 3.713147410358566,
+      "grad_norm": 0.509718120098114,
+      "learning_rate": 5.648358654324119e-07,
+      "loss": 0.9597897529602051,
+      "step": 932
+    },
+    {
+      "epoch": 3.721115537848606,
+      "grad_norm": 1.0795882940292358,
+      "learning_rate": 5.613478948473166e-07,
+      "loss": 0.5279070734977722,
+      "step": 934
+    },
+    {
+      "epoch": 3.7290836653386457,
+      "grad_norm": 0.48435333371162415,
+      "learning_rate": 5.579550611842395e-07,
+      "loss": 0.8820955753326416,
+      "step": 936
+    },
+    {
+      "epoch": 3.7370517928286855,
+      "grad_norm": 1.4125467538833618,
+      "learning_rate": 5.546575119237923e-07,
+      "loss": 0.681976318359375,
+      "step": 938
+    },
+    {
+      "epoch": 3.7450199203187253,
+      "grad_norm": 3.6143009662628174,
+      "learning_rate": 5.514553904047394e-07,
+      "loss": 0.7970798015594482,
+      "step": 940
+    },
+    {
+      "epoch": 3.752988047808765,
+      "grad_norm": 1.5276331901550293,
+      "learning_rate": 5.483488358177647e-07,
+      "loss": 0.8218612670898438,
+      "step": 942
+    },
+    {
+      "epoch": 3.760956175298805,
+      "grad_norm": 0.41266992688179016,
+      "learning_rate": 5.453379831994238e-07,
+      "loss": 0.547317624092102,
+      "step": 944
+    },
+    {
+      "epoch": 3.768924302788845,
+      "grad_norm": 0.5574037432670593,
+      "learning_rate": 5.424229634262735e-07,
+      "loss": 0.9840717315673828,
+      "step": 946
+    },
+    {
+      "epoch": 3.7768924302788847,
+      "grad_norm": 4.027848720550537,
+      "learning_rate": 5.396039032091826e-07,
+      "loss": 0.5716748237609863,
+      "step": 948
+    },
+    {
+      "epoch": 3.7848605577689245,
+      "grad_norm": 0.41914311051368713,
+      "learning_rate": 5.368809250878228e-07,
+      "loss": 0.9589597582817078,
+      "step": 950
+    },
+    {
+      "epoch": 3.7928286852589643,
+      "grad_norm": 0.7084989547729492,
+      "learning_rate": 5.342541474253438e-07,
+      "loss": 0.9919804334640503,
+      "step": 952
+    },
+    {
+      "epoch": 3.800796812749004,
+      "grad_norm": 0.5608367323875427,
+      "learning_rate": 5.317236844032285e-07,
+      "loss": 0.5673797726631165,
+      "step": 954
+    },
+    {
+      "epoch": 3.808764940239044,
+      "grad_norm": 1.0341672897338867,
+      "learning_rate": 5.292896460163282e-07,
+      "loss": 0.8357974290847778,
+      "step": 956
+    },
+    {
+      "epoch": 3.816733067729084,
+      "grad_norm": 3.5178771018981934,
+      "learning_rate": 5.269521380680823e-07,
+      "loss": 0.5885995626449585,
+      "step": 958
+    },
+    {
+      "epoch": 3.8247011952191237,
+      "grad_norm": 0.8585947155952454,
+      "learning_rate": 5.247112621659186e-07,
+      "loss": 0.7343383431434631,
+      "step": 960
+    },
+    {
+      "epoch": 3.8326693227091635,
+      "grad_norm": 0.24880963563919067,
+      "learning_rate": 5.225671157168378e-07,
+      "loss": 0.5606006979942322,
+      "step": 962
+    },
+    {
+      "epoch": 3.8406374501992033,
+      "grad_norm": 0.942073404788971,
+      "learning_rate": 5.205197919231773e-07,
+      "loss": 1.0028009414672852,
+      "step": 964
+    },
+    {
+      "epoch": 3.848605577689243,
+      "grad_norm": 0.7521424889564514,
+      "learning_rate": 5.185693797785624e-07,
+      "loss": 0.41146183013916016,
+      "step": 966
+    },
+    {
+      "epoch": 3.856573705179283,
+      "grad_norm": 0.5941879153251648,
+      "learning_rate": 5.167159640640358e-07,
+      "loss": 0.5684161186218262,
+      "step": 968
+    },
+    {
+      "epoch": 3.864541832669323,
+      "grad_norm": 1.6082491874694824,
+      "learning_rate": 5.149596253443733e-07,
+      "loss": 0.7479365468025208,
+      "step": 970
+    },
+    {
+      "epoch": 3.8725099601593627,
+      "grad_norm": 1.3756325244903564,
+      "learning_rate": 5.133004399645818e-07,
+      "loss": 0.5629640221595764,
+      "step": 972
+    },
+    {
+      "epoch": 3.8804780876494025,
+      "grad_norm": 0.1556144505739212,
+      "learning_rate": 5.117384800465806e-07,
+      "loss": 0.4646053612232208,
+      "step": 974
+    },
+    {
+      "epoch": 3.8884462151394423,
+      "grad_norm": 1.4110411405563354,
+      "learning_rate": 5.102738134860656e-07,
+      "loss": 0.9635608196258545,
+      "step": 976
+    },
+    {
+      "epoch": 3.896414342629482,
+      "grad_norm": 0.42341679334640503,
+      "learning_rate": 5.089065039495598e-07,
+      "loss": 0.5532403588294983,
+      "step": 978
+    },
+    {
+      "epoch": 3.904382470119522,
+      "grad_norm": 0.22362415492534637,
+      "learning_rate": 5.076366108716438e-07,
+      "loss": 0.3846767246723175,
+      "step": 980
+    },
+    {
+      "epoch": 3.912350597609562,
+      "grad_norm": 0.6488380432128906,
+      "learning_rate": 5.064641894523737e-07,
+      "loss": 0.4865783452987671,
+      "step": 982
+    },
+    {
+      "epoch": 3.9203187250996017,
+      "grad_norm": 0.6956201791763306,
+      "learning_rate": 5.053892906548811e-07,
+      "loss": 0.9066143035888672,
+      "step": 984
+    },
+    {
+      "epoch": 3.9282868525896415,
+      "grad_norm": 0.5712003111839294,
+      "learning_rate": 5.044119612031576e-07,
+      "loss": 0.1715915948152542,
+      "step": 986
+    },
+    {
+      "epoch": 3.9362549800796813,
+      "grad_norm": 2.1550028324127197,
+      "learning_rate": 5.035322435800248e-07,
+      "loss": 0.6000520586967468,
+      "step": 988
+    },
+    {
+      "epoch": 3.944223107569721,
+      "grad_norm": 0.1695941537618637,
+      "learning_rate": 5.027501760252857e-07,
+      "loss": 0.5861656665802002,
+      "step": 990
+    },
+    {
+      "epoch": 3.952191235059761,
+      "grad_norm": 1.306004285812378,
+      "learning_rate": 5.02065792534065e-07,
+      "loss": 1.0388450622558594,
+      "step": 992
+    },
+    {
+      "epoch": 3.960159362549801,
+      "grad_norm": 0.5286323428153992,
+      "learning_rate": 5.014791228553294e-07,
+      "loss": 0.2825939655303955,
+      "step": 994
+    },
+    {
+      "epoch": 3.9681274900398407,
+      "grad_norm": 0.6327877640724182,
+      "learning_rate": 5.00990192490595e-07,
+      "loss": 0.89066082239151,
+      "step": 996
+    },
+    {
+      "epoch": 3.9760956175298805,
+      "grad_norm": 0.16942134499549866,
+      "learning_rate": 5.005990226928196e-07,
+      "loss": 0.5648642182350159,
+      "step": 998
+    },
+    {
+      "epoch": 3.9840637450199203,
+      "grad_norm": 1.369153380393982,
+      "learning_rate": 5.003056304654774e-07,
+      "loss": 0.4663788378238678,
+      "step": 1000
+    },
+    {
+      "epoch": 3.99203187250996,
+      "grad_norm": 0.5031861066818237,
+      "learning_rate": 5.001100285618214e-07,
+      "loss": 0.9673660397529602,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.31397780776023865,
+      "learning_rate": 5.000122254843282e-07,
+      "loss": 0.35781651735305786,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0,
+      "step": 1004,
+      "total_flos": 3.519329211045118e+18,
+      "train_loss": 0.882850165816536,
+      "train_runtime": 9006.0296,
+      "train_samples_per_second": 6.689,
+      "train_steps_per_second": 0.111
+    }
+  ],
+  "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