Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

.ipynb_checkpoints/README-checkpoint.md

@@ -0,0 +1,63 @@
+---
+library_name: transformers
+license: other
+base_model: Qwen3.5-9B
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: qwen35_caption_galore
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# qwen35_caption_galore
+
+This model is a fine-tuned version of [/workspace/models/Qwen3.5-9B](https://huggingface.co//workspace/models/Qwen3.5-9B) on the my_caption dataset.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+-  family_to_adamw_lr = {
+        "language": _fallback(getattr(training_args, "language_adamw_lr", 1e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_adamw_lr", 1e-5), 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 +1,63 @@
+---
+library_name: transformers
+license: other
+base_model: Qwen3.5-9B
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: qwen35_caption_galore
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# qwen35_caption_galore
+
+This model is a fine-tuned version of [/workspace/models/Qwen3.5-9B](https://huggingface.co//workspace/models/Qwen3.5-9B) on the my_caption dataset.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+-  family_to_adamw_lr = {
+        "language": _fallback(getattr(training_args, "language_adamw_lr", 1e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_adamw_lr", 1e-5), 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": 7734.120440513665,
+    "epoch": 4.0,
+    "total_flos": 4.038502240003031e+18,
+    "train_loss": 0.8272027168051893,
+    "train_runtime": 10128.8027,
+    "train_samples_per_second": 5.947,
+    "train_steps_per_second": 0.099
+}

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 sha256:23d73d166d3be0e0eb4019ac6f3f1249bc9453a6f7a5597a21628c6fde63ccc3
+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": 7734.120440513665,
+    "epoch": 4.0,
+    "total_flos": 4.038502240003031e+18,
+    "train_loss": 0.8272027168051893,
+    "train_runtime": 10128.8027,
+    "train_samples_per_second": 5.947,
+    "train_steps_per_second": 0.099
+}

trainer_state.json

@@ -0,0 +1,3557 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 500,
+  "global_step": 1004,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00796812749003984,
+      "grad_norm": 0.9487342834472656,
+      "learning_rate": 1.9607843137254904e-07,
+      "loss": 2.0917508602142334,
+      "step": 2
+    },
+    {
+      "epoch": 0.01593625498007968,
+      "grad_norm": 0.74760502576828,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 2.0095996856689453,
+      "step": 4
+    },
+    {
+      "epoch": 0.02390438247011952,
+      "grad_norm": 5.317529201507568,
+      "learning_rate": 9.80392156862745e-07,
+      "loss": 2.184152364730835,
+      "step": 6
+    },
+    {
+      "epoch": 0.03187250996015936,
+      "grad_norm": 0.8170209527015686,
+      "learning_rate": 1.3725490196078434e-06,
+      "loss": 1.8896074295043945,
+      "step": 8
+    },
+    {
+      "epoch": 0.0398406374501992,
+      "grad_norm": 3.0944557189941406,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 1.4700803756713867,
+      "step": 10
+    },
+    {
+      "epoch": 0.04780876494023904,
+      "grad_norm": 1.2149711847305298,
+      "learning_rate": 2.1568627450980393e-06,
+      "loss": 1.7030103206634521,
+      "step": 12
+    },
+    {
+      "epoch": 0.055776892430278883,
+      "grad_norm": 0.5559038519859314,
+      "learning_rate": 2.549019607843137e-06,
+      "loss": 2.2877771854400635,
+      "step": 14
+    },
+    {
+      "epoch": 0.06374501992031872,
+      "grad_norm": 2.9151430130004883,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 2.249234199523926,
+      "step": 16
+    },
+    {
+      "epoch": 0.07171314741035857,
+      "grad_norm": 0.8587057590484619,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.6161576509475708,
+      "step": 18
+    },
+    {
+      "epoch": 0.0796812749003984,
+      "grad_norm": 0.6669076681137085,
+      "learning_rate": 3.7254901960784316e-06,
+      "loss": 1.5449607372283936,
+      "step": 20
+    },
+    {
+      "epoch": 0.08764940239043825,
+      "grad_norm": 0.4728223383426666,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 1.235238790512085,
+      "step": 22
+    },
+    {
+      "epoch": 0.09561752988047809,
+      "grad_norm": 0.7017782926559448,
+      "learning_rate": 4.509803921568628e-06,
+      "loss": 1.3569639921188354,
+      "step": 24
+    },
+    {
+      "epoch": 0.10358565737051793,
+      "grad_norm": 0.48444798588752747,
+      "learning_rate": 4.901960784313726e-06,
+      "loss": 1.448333978652954,
+      "step": 26
+    },
+    {
+      "epoch": 0.11155378486055777,
+      "grad_norm": 0.3514111042022705,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 1.6302367448806763,
+      "step": 28
+    },
+    {
+      "epoch": 0.11952191235059761,
+      "grad_norm": 0.7964296936988831,
+      "learning_rate": 5.686274509803922e-06,
+      "loss": 1.3324204683303833,
+      "step": 30
+    },
+    {
+      "epoch": 0.12749003984063745,
+      "grad_norm": 0.7039128541946411,
+      "learning_rate": 6.07843137254902e-06,
+      "loss": 1.1615524291992188,
+      "step": 32
+    },
+    {
+      "epoch": 0.13545816733067728,
+      "grad_norm": 0.4866271913051605,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 1.2443053722381592,
+      "step": 34
+    },
+    {
+      "epoch": 0.14342629482071714,
+      "grad_norm": 0.44938722252845764,
+      "learning_rate": 6.862745098039216e-06,
+      "loss": 1.4063172340393066,
+      "step": 36
+    },
+    {
+      "epoch": 0.15139442231075698,
+      "grad_norm": 0.5627224445343018,
+      "learning_rate": 7.2549019607843145e-06,
+      "loss": 1.190273642539978,
+      "step": 38
+    },
+    {
+      "epoch": 0.1593625498007968,
+      "grad_norm": 0.6883392333984375,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 1.3222601413726807,
+      "step": 40
+    },
+    {
+      "epoch": 0.16733067729083664,
+      "grad_norm": 0.7158175706863403,
+      "learning_rate": 8.03921568627451e-06,
+      "loss": 0.7816150188446045,
+      "step": 42
+    },
+    {
+      "epoch": 0.1752988047808765,
+      "grad_norm": 0.5107442736625671,
+      "learning_rate": 8.43137254901961e-06,
+      "loss": 0.9847684502601624,
+      "step": 44
+    },
+    {
+      "epoch": 0.18326693227091634,
+      "grad_norm": 1.4377588033676147,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.9455029964447021,
+      "step": 46
+    },
+    {
+      "epoch": 0.19123505976095617,
+      "grad_norm": 0.6867519617080688,
+      "learning_rate": 9.215686274509804e-06,
+      "loss": 1.1748902797698975,
+      "step": 48
+    },
+    {
+      "epoch": 0.199203187250996,
+      "grad_norm": 0.4950691759586334,
+      "learning_rate": 9.607843137254903e-06,
+      "loss": 1.3410654067993164,
+      "step": 50
+    },
+    {
+      "epoch": 0.20717131474103587,
+      "grad_norm": 0.7239283919334412,
+      "learning_rate": 1e-05,
+      "loss": 1.2962646484375,
+      "step": 52
+    },
+    {
+      "epoch": 0.2151394422310757,
+      "grad_norm": 1.614442229270935,
+      "learning_rate": 9.999902196391087e-06,
+      "loss": 1.0902235507965088,
+      "step": 54
+    },
+    {
+      "epoch": 0.22310756972111553,
+      "grad_norm": 0.47346577048301697,
+      "learning_rate": 9.999608789815699e-06,
+      "loss": 1.5185879468917847,
+      "step": 56
+    },
+    {
+      "epoch": 0.23107569721115537,
+      "grad_norm": 0.4998139441013336,
+      "learning_rate": 9.999119793027713e-06,
+      "loss": 1.3452786207199097,
+      "step": 58
+    },
+    {
+      "epoch": 0.23904382470119523,
+      "grad_norm": 0.45596635341644287,
+      "learning_rate": 9.998435227282975e-06,
+      "loss": 0.8151512742042542,
+      "step": 60
+    },
+    {
+      "epoch": 0.24701195219123506,
+      "grad_norm": 0.4687139689922333,
+      "learning_rate": 9.997555122338372e-06,
+      "loss": 1.3474674224853516,
+      "step": 62
+    },
+    {
+      "epoch": 0.2549800796812749,
+      "grad_norm": 1.1674251556396484,
+      "learning_rate": 9.996479516450544e-06,
+      "loss": 1.1081370115280151,
+      "step": 64
+    },
+    {
+      "epoch": 0.26294820717131473,
+      "grad_norm": 0.35629168152809143,
+      "learning_rate": 9.99520845637422e-06,
+      "loss": 1.2745870351791382,
+      "step": 66
+    },
+    {
+      "epoch": 0.27091633466135456,
+      "grad_norm": 0.9294307827949524,
+      "learning_rate": 9.993741997360185e-06,
+      "loss": 0.8528879880905151,
+      "step": 68
+    },
+    {
+      "epoch": 0.2788844621513944,
+      "grad_norm": 0.8441171646118164,
+      "learning_rate": 9.992080203152872e-06,
+      "loss": 1.542641520500183,
+      "step": 70
+    },
+    {
+      "epoch": 0.2868525896414343,
+      "grad_norm": 0.7050015926361084,
+      "learning_rate": 9.990223145987608e-06,
+      "loss": 0.8513721227645874,
+      "step": 72
+    },
+    {
+      "epoch": 0.2948207171314741,
+      "grad_norm": 0.8692718148231506,
+      "learning_rate": 9.98817090658745e-06,
+      "loss": 1.1390178203582764,
+      "step": 74
+    },
+    {
+      "epoch": 0.30278884462151395,
+      "grad_norm": 0.9042101502418518,
+      "learning_rate": 9.985923574159696e-06,
+      "loss": 1.4578380584716797,
+      "step": 76
+    },
+    {
+      "epoch": 0.3107569721115538,
+      "grad_norm": 1.4772430658340454,
+      "learning_rate": 9.983481246392003e-06,
+      "loss": 0.9475153088569641,
+      "step": 78
+    },
+    {
+      "epoch": 0.3187250996015936,
+      "grad_norm": 0.38654640316963196,
+      "learning_rate": 9.980844029448135e-06,
+      "loss": 0.8545466661453247,
+      "step": 80
+    },
+    {
+      "epoch": 0.32669322709163345,
+      "grad_norm": 0.4513205587863922,
+      "learning_rate": 9.97801203796335e-06,
+      "loss": 1.3327709436416626,
+      "step": 82
+    },
+    {
+      "epoch": 0.3346613545816733,
+      "grad_norm": 1.3083326816558838,
+      "learning_rate": 9.974985395039423e-06,
+      "loss": 0.9056742191314697,
+      "step": 84
+    },
+    {
+      "epoch": 0.3426294820717131,
+      "grad_norm": 0.40296682715415955,
+      "learning_rate": 9.971764232239287e-06,
+      "loss": 1.31935453414917,
+      "step": 86
+    },
+    {
+      "epoch": 0.350597609561753,
+      "grad_norm": 5.301670551300049,
+      "learning_rate": 9.968348689581315e-06,
+      "loss": 0.9326410293579102,
+      "step": 88
+    },
+    {
+      "epoch": 0.35856573705179284,
+      "grad_norm": 1.2513582706451416,
+      "learning_rate": 9.964738915533242e-06,
+      "loss": 1.3306196928024292,
+      "step": 90
+    },
+    {
+      "epoch": 0.3665338645418327,
+      "grad_norm": 1.2525602579116821,
+      "learning_rate": 9.960935067005706e-06,
+      "loss": 1.130804419517517,
+      "step": 92
+    },
+    {
+      "epoch": 0.3745019920318725,
+      "grad_norm": 0.5879653096199036,
+      "learning_rate": 9.956937309345418e-06,
+      "loss": 1.2423148155212402,
+      "step": 94
+    },
+    {
+      "epoch": 0.38247011952191234,
+      "grad_norm": 2.2197742462158203,
+      "learning_rate": 9.952745816327998e-06,
+      "loss": 1.094862461090088,
+      "step": 96
+    },
+    {
+      "epoch": 0.3904382470119522,
+      "grad_norm": 0.35003992915153503,
+      "learning_rate": 9.94836077015039e-06,
+      "loss": 1.2666294574737549,
+      "step": 98
+    },
+    {
+      "epoch": 0.398406374501992,
+      "grad_norm": 0.4932577908039093,
+      "learning_rate": 9.943782361422975e-06,
+      "loss": 0.876234769821167,
+      "step": 100
+    },
+    {
+      "epoch": 0.4063745019920319,
+      "grad_norm": 0.6704363226890564,
+      "learning_rate": 9.939010789161259e-06,
+      "loss": 1.2801531553268433,
+      "step": 102
+    },
+    {
+      "epoch": 0.41434262948207173,
+      "grad_norm": 0.373979389667511,
+      "learning_rate": 9.934046260777239e-06,
+      "loss": 0.7953701615333557,
+      "step": 104
+    },
+    {
+      "epoch": 0.42231075697211157,
+      "grad_norm": 2.2562756538391113,
+      "learning_rate": 9.928888992070373e-06,
+      "loss": 0.7333505749702454,
+      "step": 106
+    },
+    {
+      "epoch": 0.4302788844621514,
+      "grad_norm": 0.6253886222839355,
+      "learning_rate": 9.923539207218219e-06,
+      "loss": 1.0008593797683716,
+      "step": 108
+    },
+    {
+      "epoch": 0.43824701195219123,
+      "grad_norm": 0.48899900913238525,
+      "learning_rate": 9.91799713876667e-06,
+      "loss": 1.299308180809021,
+      "step": 110
+    },
+    {
+      "epoch": 0.44621513944223107,
+      "grad_norm": 0.6760080456733704,
+      "learning_rate": 9.912263027619856e-06,
+      "loss": 0.845918595790863,
+      "step": 112
+    },
+    {
+      "epoch": 0.4541832669322709,
+      "grad_norm": 0.5138827562332153,
+      "learning_rate": 9.906337123029675e-06,
+      "loss": 1.1567695140838623,
+      "step": 114
+    },
+    {
+      "epoch": 0.46215139442231074,
+      "grad_norm": 0.6125851273536682,
+      "learning_rate": 9.90021968258495e-06,
+      "loss": 1.3113179206848145,
+      "step": 116
+    },
+    {
+      "epoch": 0.4701195219123506,
+      "grad_norm": 0.4265132248401642,
+      "learning_rate": 9.893910972200238e-06,
+      "loss": 1.3601348400115967,
+      "step": 118
+    },
+    {
+      "epoch": 0.47808764940239046,
+      "grad_norm": 0.6269770860671997,
+      "learning_rate": 9.887411266104268e-06,
+      "loss": 1.245082139968872,
+      "step": 120
+    },
+    {
+      "epoch": 0.4860557768924303,
+      "grad_norm": 0.3228493332862854,
+      "learning_rate": 9.880720846828026e-06,
+      "loss": 0.9631037712097168,
+      "step": 122
+    },
+    {
+      "epoch": 0.4940239043824701,
+      "grad_norm": 1.1927493810653687,
+      "learning_rate": 9.873840005192466e-06,
+      "loss": 1.0419517755508423,
+      "step": 124
+    },
+    {
+      "epoch": 0.50199203187251,
+      "grad_norm": 0.7198822498321533,
+      "learning_rate": 9.866769040295875e-06,
+      "loss": 0.934771716594696,
+      "step": 126
+    },
+    {
+      "epoch": 0.5099601593625498,
+      "grad_norm": 0.412668913602829,
+      "learning_rate": 9.859508259500869e-06,
+      "loss": 1.284218430519104,
+      "step": 128
+    },
+    {
+      "epoch": 0.5179282868525896,
+      "grad_norm": 0.45108819007873535,
+      "learning_rate": 9.852057978421033e-06,
+      "loss": 0.9475380778312683,
+      "step": 130
+    },
+    {
+      "epoch": 0.5258964143426295,
+      "grad_norm": 1.0086629390716553,
+      "learning_rate": 9.844418520907198e-06,
+      "loss": 0.8096392154693604,
+      "step": 132
+    },
+    {
+      "epoch": 0.5338645418326693,
+      "grad_norm": 0.4596579074859619,
+      "learning_rate": 9.836590219033375e-06,
+      "loss": 0.908939003944397,
+      "step": 134
+    },
+    {
+      "epoch": 0.5418326693227091,
+      "grad_norm": 1.650301456451416,
+      "learning_rate": 9.828573413082306e-06,
+      "loss": 1.284470796585083,
+      "step": 136
+    },
+    {
+      "epoch": 0.549800796812749,
+      "grad_norm": 0.37744879722595215,
+      "learning_rate": 9.820368451530681e-06,
+      "loss": 0.9610538482666016,
+      "step": 138
+    },
+    {
+      "epoch": 0.5577689243027888,
+      "grad_norm": 1.1511077880859375,
+      "learning_rate": 9.811975691033992e-06,
+      "loss": 1.0894925594329834,
+      "step": 140
+    },
+    {
+      "epoch": 0.5657370517928287,
+      "grad_norm": 0.6419306993484497,
+      "learning_rate": 9.80339549641102e-06,
+      "loss": 1.0302261114120483,
+      "step": 142
+    },
+    {
+      "epoch": 0.5737051792828686,
+      "grad_norm": 2.7169275283813477,
+      "learning_rate": 9.794628240627992e-06,
+      "loss": 1.236246943473816,
+      "step": 144
+    },
+    {
+      "epoch": 0.5816733067729084,
+      "grad_norm": 0.8609069585800171,
+      "learning_rate": 9.785674304782349e-06,
+      "loss": 1.2335845232009888,
+      "step": 146
+    },
+    {
+      "epoch": 0.5896414342629482,
+      "grad_norm": 0.7360414862632751,
+      "learning_rate": 9.776534078086202e-06,
+      "loss": 0.9768642783164978,
+      "step": 148
+    },
+    {
+      "epoch": 0.5976095617529881,
+      "grad_norm": 0.8006039261817932,
+      "learning_rate": 9.767207957849396e-06,
+      "loss": 0.9541663527488708,
+      "step": 150
+    },
+    {
+      "epoch": 0.6055776892430279,
+      "grad_norm": 0.7450659275054932,
+      "learning_rate": 9.757696349462252e-06,
+      "loss": 0.9391952157020569,
+      "step": 152
+    },
+    {
+      "epoch": 0.6135458167330677,
+      "grad_norm": 2.2794835567474365,
+      "learning_rate": 9.747999666377934e-06,
+      "loss": 1.12716805934906,
+      "step": 154
+    },
+    {
+      "epoch": 0.6215139442231076,
+      "grad_norm": 4.2820940017700195,
+      "learning_rate": 9.738118330094491e-06,
+      "loss": 1.0286604166030884,
+      "step": 156
+    },
+    {
+      "epoch": 0.6294820717131474,
+      "grad_norm": 1.9364804029464722,
+      "learning_rate": 9.728052770136518e-06,
+      "loss": 1.1623810529708862,
+      "step": 158
+    },
+    {
+      "epoch": 0.6374501992031872,
+      "grad_norm": 0.45877596735954285,
+      "learning_rate": 9.717803424036501e-06,
+      "loss": 1.0298248529434204,
+      "step": 160
+    },
+    {
+      "epoch": 0.6454183266932271,
+      "grad_norm": 0.2666255235671997,
+      "learning_rate": 9.707370737315793e-06,
+      "loss": 0.7623182535171509,
+      "step": 162
+    },
+    {
+      "epoch": 0.6533864541832669,
+      "grad_norm": 0.335033655166626,
+      "learning_rate": 9.69675516346524e-06,
+      "loss": 1.167417287826538,
+      "step": 164
+    },
+    {
+      "epoch": 0.6613545816733067,
+      "grad_norm": 0.5316612720489502,
+      "learning_rate": 9.685957163925479e-06,
+      "loss": 0.82045578956604,
+      "step": 166
+    },
+    {
+      "epoch": 0.6693227091633466,
+      "grad_norm": 0.6057591438293457,
+      "learning_rate": 9.67497720806688e-06,
+      "loss": 1.2575392723083496,
+      "step": 168
+    },
+    {
+      "epoch": 0.6772908366533864,
+      "grad_norm": 2.0181779861450195,
+      "learning_rate": 9.663815773169131e-06,
+      "loss": 0.9410154223442078,
+      "step": 170
+    },
+    {
+      "epoch": 0.6852589641434262,
+      "grad_norm": 0.511999249458313,
+      "learning_rate": 9.652473344400507e-06,
+      "loss": 0.8805829286575317,
+      "step": 172
+    },
+    {
+      "epoch": 0.6932270916334662,
+      "grad_norm": 0.422842800617218,
+      "learning_rate": 9.640950414796772e-06,
+      "loss": 1.248655915260315,
+      "step": 174
+    },
+    {
+      "epoch": 0.701195219123506,
+      "grad_norm": 1.3418399095535278,
+      "learning_rate": 9.629247485239747e-06,
+      "loss": 0.7680034041404724,
+      "step": 176
+    },
+    {
+      "epoch": 0.7091633466135459,
+      "grad_norm": 1.0295521020889282,
+      "learning_rate": 9.617365064435544e-06,
+      "loss": 0.9724248647689819,
+      "step": 178
+    },
+    {
+      "epoch": 0.7171314741035857,
+      "grad_norm": 0.7610090970993042,
+      "learning_rate": 9.605303668892443e-06,
+      "loss": 1.0979655981063843,
+      "step": 180
+    },
+    {
+      "epoch": 0.7250996015936255,
+      "grad_norm": 0.29494383931159973,
+      "learning_rate": 9.593063822898451e-06,
+      "loss": 0.5257850289344788,
+      "step": 182
+    },
+    {
+      "epoch": 0.7330677290836654,
+      "grad_norm": 0.36800092458724976,
+      "learning_rate": 9.580646058498505e-06,
+      "loss": 1.3627965450286865,
+      "step": 184
+    },
+    {
+      "epoch": 0.7410358565737052,
+      "grad_norm": 0.4371291697025299,
+      "learning_rate": 9.56805091547135e-06,
+      "loss": 1.3105531930923462,
+      "step": 186
+    },
+    {
+      "epoch": 0.749003984063745,
+      "grad_norm": 0.5148975849151611,
+      "learning_rate": 9.55527894130607e-06,
+      "loss": 1.1319502592086792,
+      "step": 188
+    },
+    {
+      "epoch": 0.7569721115537849,
+      "grad_norm": 0.42739081382751465,
+      "learning_rate": 9.542330691178296e-06,
+      "loss": 0.8949350118637085,
+      "step": 190
+    },
+    {
+      "epoch": 0.7649402390438247,
+      "grad_norm": 0.732265055179596,
+      "learning_rate": 9.529206727926071e-06,
+      "loss": 1.247640609741211,
+      "step": 192
+    },
+    {
+      "epoch": 0.7729083665338645,
+      "grad_norm": 0.562000572681427,
+      "learning_rate": 9.515907622025376e-06,
+      "loss": 0.9021010994911194,
+      "step": 194
+    },
+    {
+      "epoch": 0.7808764940239044,
+      "grad_norm": 0.6893383860588074,
+      "learning_rate": 9.502433951565347e-06,
+      "loss": 0.5976129174232483,
+      "step": 196
+    },
+    {
+      "epoch": 0.7888446215139442,
+      "grad_norm": 0.8131871819496155,
+      "learning_rate": 9.48878630222314e-06,
+      "loss": 0.801398515701294,
+      "step": 198
+    },
+    {
+      "epoch": 0.796812749003984,
+      "grad_norm": 0.37410953640937805,
+      "learning_rate": 9.474965267238469e-06,
+      "loss": 1.072272539138794,
+      "step": 200
+    },
+    {
+      "epoch": 0.8047808764940239,
+      "grad_norm": 0.19710387289524078,
+      "learning_rate": 9.460971447387824e-06,
+      "loss": 0.987641453742981,
+      "step": 202
+    },
+    {
+      "epoch": 0.8127490039840638,
+      "grad_norm": 0.4658758044242859,
+      "learning_rate": 9.446805450958353e-06,
+      "loss": 1.5645253658294678,
+      "step": 204
+    },
+    {
+      "epoch": 0.8207171314741036,
+      "grad_norm": 0.3566979169845581,
+      "learning_rate": 9.43246789372142e-06,
+      "loss": 1.2662503719329834,
+      "step": 206
+    },
+    {
+      "epoch": 0.8286852589641435,
+      "grad_norm": 0.2858918011188507,
+      "learning_rate": 9.417959398905846e-06,
+      "loss": 0.6095532178878784,
+      "step": 208
+    },
+    {
+      "epoch": 0.8366533864541833,
+      "grad_norm": 0.33042585849761963,
+      "learning_rate": 9.40328059717081e-06,
+      "loss": 1.3066667318344116,
+      "step": 210
+    },
+    {
+      "epoch": 0.8446215139442231,
+      "grad_norm": 0.660761296749115,
+      "learning_rate": 9.388432126578438e-06,
+      "loss": 1.2502214908599854,
+      "step": 212
+    },
+    {
+      "epoch": 0.852589641434263,
+      "grad_norm": 0.3721761405467987,
+      "learning_rate": 9.373414632566066e-06,
+      "loss": 0.9593994617462158,
+      "step": 214
+    },
+    {
+      "epoch": 0.8605577689243028,
+      "grad_norm": 1.7799036502838135,
+      "learning_rate": 9.35822876791819e-06,
+      "loss": 1.4516774415969849,
+      "step": 216
+    },
+    {
+      "epoch": 0.8685258964143426,
+      "grad_norm": 0.9014025330543518,
+      "learning_rate": 9.342875192738083e-06,
+      "loss": 0.8795238733291626,
+      "step": 218
+    },
+    {
+      "epoch": 0.8764940239043825,
+      "grad_norm": 0.29185619950294495,
+      "learning_rate": 9.327354574419109e-06,
+      "loss": 1.2557966709136963,
+      "step": 220
+    },
+    {
+      "epoch": 0.8844621513944223,
+      "grad_norm": 4.0524396896362305,
+      "learning_rate": 9.311667587615701e-06,
+      "loss": 1.5041028261184692,
+      "step": 222
+    },
+    {
+      "epoch": 0.8924302788844621,
+      "grad_norm": 0.6374480128288269,
+      "learning_rate": 9.29581491421405e-06,
+      "loss": 0.7121538519859314,
+      "step": 224
+    },
+    {
+      "epoch": 0.900398406374502,
+      "grad_norm": 0.33715274930000305,
+      "learning_rate": 9.279797243302453e-06,
+      "loss": 1.0032306909561157,
+      "step": 226
+    },
+    {
+      "epoch": 0.9083665338645418,
+      "grad_norm": 0.8425693511962891,
+      "learning_rate": 9.263615271141362e-06,
+      "loss": 1.1027175188064575,
+      "step": 228
+    },
+    {
+      "epoch": 0.9163346613545816,
+      "grad_norm": 0.5570669174194336,
+      "learning_rate": 9.247269701133125e-06,
+      "loss": 1.1932541131973267,
+      "step": 230
+    },
+    {
+      "epoch": 0.9243027888446215,
+      "grad_norm": 0.3410831689834595,
+      "learning_rate": 9.230761243791401e-06,
+      "loss": 1.3409861326217651,
+      "step": 232
+    },
+    {
+      "epoch": 0.9322709163346613,
+      "grad_norm": 0.3947798013687134,
+      "learning_rate": 9.214090616710282e-06,
+      "loss": 0.8251549005508423,
+      "step": 234
+    },
+    {
+      "epoch": 0.9402390438247012,
+      "grad_norm": 0.4384807050228119,
+      "learning_rate": 9.197258544533098e-06,
+      "loss": 0.8354034423828125,
+      "step": 236
+    },
+    {
+      "epoch": 0.9482071713147411,
+      "grad_norm": 1.005926489830017,
+      "learning_rate": 9.180265758920916e-06,
+      "loss": 0.5023077130317688,
+      "step": 238
+    },
+    {
+      "epoch": 0.9561752988047809,
+      "grad_norm": 0.7860603332519531,
+      "learning_rate": 9.163112998520741e-06,
+      "loss": 0.8660817742347717,
+      "step": 240
+    },
+    {
+      "epoch": 0.9641434262948207,
+      "grad_norm": 0.41115736961364746,
+      "learning_rate": 9.145801008933403e-06,
+      "loss": 1.148357629776001,
+      "step": 242
+    },
+    {
+      "epoch": 0.9721115537848606,
+      "grad_norm": 0.26804134249687195,
+      "learning_rate": 9.128330542681154e-06,
+      "loss": 1.2330094575881958,
+      "step": 244
+    },
+    {
+      "epoch": 0.9800796812749004,
+      "grad_norm": 0.6420534253120422,
+      "learning_rate": 9.110702359174944e-06,
+      "loss": 0.9773095846176147,
+      "step": 246
+    },
+    {
+      "epoch": 0.9880478087649402,
+      "grad_norm": 0.29611170291900635,
+      "learning_rate": 9.092917224681427e-06,
+      "loss": 1.0235339403152466,
+      "step": 248
+    },
+    {
+      "epoch": 0.9960159362549801,
+      "grad_norm": 1.1437735557556152,
+      "learning_rate": 9.074975912289641e-06,
+      "loss": 1.001664400100708,
+      "step": 250
+    },
+    {
+      "epoch": 1.00398406374502,
+      "grad_norm": 0.6782959699630737,
+      "learning_rate": 9.056879201877412e-06,
+      "loss": 0.9294381141662598,
+      "step": 252
+    },
+    {
+      "epoch": 1.0119521912350598,
+      "grad_norm": 0.23827040195465088,
+      "learning_rate": 9.038627880077441e-06,
+      "loss": 0.6967454552650452,
+      "step": 254
+    },
+    {
+      "epoch": 1.0199203187250996,
+      "grad_norm": 0.40992769598960876,
+      "learning_rate": 9.020222740243127e-06,
+      "loss": 1.146529197692871,
+      "step": 256
+    },
+    {
+      "epoch": 1.0278884462151394,
+      "grad_norm": 0.3573630750179291,
+      "learning_rate": 9.001664582414066e-06,
+      "loss": 1.1288363933563232,
+      "step": 258
+    },
+    {
+      "epoch": 1.0358565737051793,
+      "grad_norm": 0.5140087604522705,
+      "learning_rate": 8.982954213281291e-06,
+      "loss": 1.0338488817214966,
+      "step": 260
+    },
+    {
+      "epoch": 1.043824701195219,
+      "grad_norm": 0.3772323429584503,
+      "learning_rate": 8.964092446152183e-06,
+      "loss": 1.0023149251937866,
+      "step": 262
+    },
+    {
+      "epoch": 1.051792828685259,
+      "grad_norm": 0.5765355229377747,
+      "learning_rate": 8.945080100915145e-06,
+      "loss": 1.048535704612732,
+      "step": 264
+    },
+    {
+      "epoch": 1.0597609561752988,
+      "grad_norm": 0.6774141192436218,
+      "learning_rate": 8.925918004003941e-06,
+      "loss": 0.6529957056045532,
+      "step": 266
+    },
+    {
+      "epoch": 1.0677290836653386,
+      "grad_norm": 0.27868425846099854,
+      "learning_rate": 8.90660698836179e-06,
+      "loss": 0.8979067206382751,
+      "step": 268
+    },
+    {
+      "epoch": 1.0756972111553784,
+      "grad_norm": 0.35501497983932495,
+      "learning_rate": 8.887147893405145e-06,
+      "loss": 0.555001437664032,
+      "step": 270
+    },
+    {
+      "epoch": 1.0836653386454183,
+      "grad_norm": 0.3792285919189453,
+      "learning_rate": 8.867541564987212e-06,
+      "loss": 1.2294905185699463,
+      "step": 272
+    },
+    {
+      "epoch": 1.091633466135458,
+      "grad_norm": 0.4947018623352051,
+      "learning_rate": 8.84778885536118e-06,
+      "loss": 0.6028950810432434,
+      "step": 274
+    },
+    {
+      "epoch": 1.099601593625498,
+      "grad_norm": 0.41941797733306885,
+      "learning_rate": 8.827890623143172e-06,
+      "loss": 0.8728867173194885,
+      "step": 276
+    },
+    {
+      "epoch": 1.1075697211155378,
+      "grad_norm": 0.6124916076660156,
+      "learning_rate": 8.80784773327494e-06,
+      "loss": 0.7639375329017639,
+      "step": 278
+    },
+    {
+      "epoch": 1.1155378486055776,
+      "grad_norm": 0.3823148310184479,
+      "learning_rate": 8.787661056986236e-06,
+      "loss": 0.9852554202079773,
+      "step": 280
+    },
+    {
+      "epoch": 1.1235059760956174,
+      "grad_norm": 0.2765064239501953,
+      "learning_rate": 8.767331471756975e-06,
+      "loss": 1.168877363204956,
+      "step": 282
+    },
+    {
+      "epoch": 1.1314741035856573,
+      "grad_norm": 1.2946758270263672,
+      "learning_rate": 8.746859861279068e-06,
+      "loss": 1.1658602952957153,
+      "step": 284
+    },
+    {
+      "epoch": 1.139442231075697,
+      "grad_norm": 0.33701643347740173,
+      "learning_rate": 8.726247115418017e-06,
+      "loss": 0.27370989322662354,
+      "step": 286
+    },
+    {
+      "epoch": 1.1474103585657371,
+      "grad_norm": 0.4078567624092102,
+      "learning_rate": 8.70549413017424e-06,
+      "loss": 0.6657518148422241,
+      "step": 288
+    },
+    {
+      "epoch": 1.155378486055777,
+      "grad_norm": 0.8731371164321899,
+      "learning_rate": 8.684601807644114e-06,
+      "loss": 0.8997029066085815,
+      "step": 290
+    },
+    {
+      "epoch": 1.1633466135458168,
+      "grad_norm": 0.40931236743927,
+      "learning_rate": 8.663571055980769e-06,
+      "loss": 0.5055676102638245,
+      "step": 292
+    },
+    {
+      "epoch": 1.1713147410358566,
+      "grad_norm": 0.6988786458969116,
+      "learning_rate": 8.642402789354609e-06,
+      "loss": 1.1833248138427734,
+      "step": 294
+    },
+    {
+      "epoch": 1.1792828685258965,
+      "grad_norm": 0.25156351923942566,
+      "learning_rate": 8.62109792791358e-06,
+      "loss": 1.225422739982605,
+      "step": 296
+    },
+    {
+      "epoch": 1.1872509960159363,
+      "grad_norm": 0.3821275532245636,
+      "learning_rate": 8.599657397743163e-06,
+      "loss": 1.3129684925079346,
+      "step": 298
+    },
+    {
+      "epoch": 1.1952191235059761,
+      "grad_norm": 0.44987067580223083,
+      "learning_rate": 8.578082130826123e-06,
+      "loss": 0.6042513847351074,
+      "step": 300
+    },
+    {
+      "epoch": 1.203187250996016,
+      "grad_norm": 0.32683783769607544,
+      "learning_rate": 8.556373065002006e-06,
+      "loss": 0.9305551052093506,
+      "step": 302
+    },
+    {
+      "epoch": 1.2111553784860558,
+      "grad_norm": 0.5258086919784546,
+      "learning_rate": 8.534531143926359e-06,
+      "loss": 0.5298245549201965,
+      "step": 304
+    },
+    {
+      "epoch": 1.2191235059760956,
+      "grad_norm": 0.3306218981742859,
+      "learning_rate": 8.512557317029715e-06,
+      "loss": 1.1828659772872925,
+      "step": 306
+    },
+    {
+      "epoch": 1.2270916334661355,
+      "grad_norm": 0.9576687812805176,
+      "learning_rate": 8.49045253947633e-06,
+      "loss": 1.01560640335083,
+      "step": 308
+    },
+    {
+      "epoch": 1.2350597609561753,
+      "grad_norm": 0.5489612817764282,
+      "learning_rate": 8.46821777212266e-06,
+      "loss": 1.1288374662399292,
+      "step": 310
+    },
+    {
+      "epoch": 1.2430278884462151,
+      "grad_norm": 0.38077735900878906,
+      "learning_rate": 8.445853981475587e-06,
+      "loss": 1.0438995361328125,
+      "step": 312
+    },
+    {
+      "epoch": 1.250996015936255,
+      "grad_norm": 0.44545653462409973,
+      "learning_rate": 8.423362139650414e-06,
+      "loss": 0.8773861527442932,
+      "step": 314
+    },
+    {
+      "epoch": 1.2589641434262948,
+      "grad_norm": 0.41514939069747925,
+      "learning_rate": 8.40074322432861e-06,
+      "loss": 0.7363654375076294,
+      "step": 316
+    },
+    {
+      "epoch": 1.2669322709163346,
+      "grad_norm": 0.746155858039856,
+      "learning_rate": 8.377998218715311e-06,
+      "loss": 0.6529978513717651,
+      "step": 318
+    },
+    {
+      "epoch": 1.2749003984063745,
+      "grad_norm": 0.6861006617546082,
+      "learning_rate": 8.355128111496569e-06,
+      "loss": 0.521120548248291,
+      "step": 320
+    },
+    {
+      "epoch": 1.2828685258964143,
+      "grad_norm": 0.4531482756137848,
+      "learning_rate": 8.332133896796399e-06,
+      "loss": 0.9750446677207947,
+      "step": 322
+    },
+    {
+      "epoch": 1.2908366533864541,
+      "grad_norm": 0.7975627779960632,
+      "learning_rate": 8.309016574133545e-06,
+      "loss": 0.9169808626174927,
+      "step": 324
+    },
+    {
+      "epoch": 1.298804780876494,
+      "grad_norm": 0.6357133388519287,
+      "learning_rate": 8.28577714837805e-06,
+      "loss": 1.2473933696746826,
+      "step": 326
+    },
+    {
+      "epoch": 1.3067729083665338,
+      "grad_norm": 0.34495019912719727,
+      "learning_rate": 8.262416629707555e-06,
+      "loss": 0.7174408435821533,
+      "step": 328
+    },
+    {
+      "epoch": 1.3147410358565736,
+      "grad_norm": 0.39262065291404724,
+      "learning_rate": 8.238936033563415e-06,
+      "loss": 0.8348032832145691,
+      "step": 330
+    },
+    {
+      "epoch": 1.3227091633466135,
+      "grad_norm": 0.4772288203239441,
+      "learning_rate": 8.215336380606533e-06,
+      "loss": 0.9283754229545593,
+      "step": 332
+    },
+    {
+      "epoch": 1.3306772908366533,
+      "grad_norm": 0.8756493926048279,
+      "learning_rate": 8.191618696673012e-06,
+      "loss": 0.6008652448654175,
+      "step": 334
+    },
+    {
+      "epoch": 1.3386454183266931,
+      "grad_norm": 0.554428219795227,
+      "learning_rate": 8.167784012729555e-06,
+      "loss": 0.8723337054252625,
+      "step": 336
+    },
+    {
+      "epoch": 1.3466135458167332,
+      "grad_norm": 0.3797021806240082,
+      "learning_rate": 8.143833364828654e-06,
+      "loss": 1.1309664249420166,
+      "step": 338
+    },
+    {
+      "epoch": 1.354581673306773,
+      "grad_norm": 0.28092142939567566,
+      "learning_rate": 8.119767794063553e-06,
+      "loss": 1.134152889251709,
+      "step": 340
+    },
+    {
+      "epoch": 1.3625498007968129,
+      "grad_norm": 0.2888767421245575,
+      "learning_rate": 8.095588346523e-06,
+      "loss": 1.1758265495300293,
+      "step": 342
+    },
+    {
+      "epoch": 1.3705179282868527,
+      "grad_norm": 0.3067930340766907,
+      "learning_rate": 8.071296073245759e-06,
+      "loss": 1.1344200372695923,
+      "step": 344
+    },
+    {
+      "epoch": 1.3784860557768925,
+      "grad_norm": 0.8363769054412842,
+      "learning_rate": 8.046892030174938e-06,
+      "loss": 0.7353524565696716,
+      "step": 346
+    },
+    {
+      "epoch": 1.3864541832669324,
+      "grad_norm": 0.3148793578147888,
+      "learning_rate": 8.022377278112089e-06,
+      "loss": 1.163429856300354,
+      "step": 348
+    },
+    {
+      "epoch": 1.3944223107569722,
+      "grad_norm": 0.93288654088974,
+      "learning_rate": 7.997752882671088e-06,
+      "loss": 0.6754208207130432,
+      "step": 350
+    },
+    {
+      "epoch": 1.402390438247012,
+      "grad_norm": 0.5121386051177979,
+      "learning_rate": 7.97301991423182e-06,
+      "loss": 1.1746786832809448,
+      "step": 352
+    },
+    {
+      "epoch": 1.4103585657370519,
+      "grad_norm": 0.5762549042701721,
+      "learning_rate": 7.948179447893652e-06,
+      "loss": 1.1066590547561646,
+      "step": 354
+    },
+    {
+      "epoch": 1.4183266932270917,
+      "grad_norm": 0.7748022675514221,
+      "learning_rate": 7.923232563428693e-06,
+      "loss": 0.8950283527374268,
+      "step": 356
+    },
+    {
+      "epoch": 1.4262948207171315,
+      "grad_norm": 0.3753865957260132,
+      "learning_rate": 7.898180345234875e-06,
+      "loss": 1.1449472904205322,
+      "step": 358
+    },
+    {
+      "epoch": 1.4342629482071714,
+      "grad_norm": 0.3016361892223358,
+      "learning_rate": 7.873023882288793e-06,
+      "loss": 0.5846461653709412,
+      "step": 360
+    },
+    {
+      "epoch": 1.4422310756972112,
+      "grad_norm": 0.23003606498241425,
+      "learning_rate": 7.847764268098387e-06,
+      "loss": 0.6807458400726318,
+      "step": 362
+    },
+    {
+      "epoch": 1.450199203187251,
+      "grad_norm": 0.4251004159450531,
+      "learning_rate": 7.822402600655405e-06,
+      "loss": 1.045509934425354,
+      "step": 364
+    },
+    {
+      "epoch": 1.4581673306772909,
+      "grad_norm": 0.23218894004821777,
+      "learning_rate": 7.79693998238767e-06,
+      "loss": 1.1032710075378418,
+      "step": 366
+    },
+    {
+      "epoch": 1.4661354581673307,
+      "grad_norm": 0.23951773345470428,
+      "learning_rate": 7.771377520111163e-06,
+      "loss": 0.6049657464027405,
+      "step": 368
+    },
+    {
+      "epoch": 1.4741035856573705,
+      "grad_norm": 0.6444947719573975,
+      "learning_rate": 7.745716324981912e-06,
+      "loss": 0.8669015169143677,
+      "step": 370
+    },
+    {
+      "epoch": 1.4820717131474104,
+      "grad_norm": 0.567205011844635,
+      "learning_rate": 7.71995751244769e-06,
+      "loss": 1.2079463005065918,
+      "step": 372
+    },
+    {
+      "epoch": 1.4900398406374502,
+      "grad_norm": 0.5559811592102051,
+      "learning_rate": 7.694102202199536e-06,
+      "loss": 1.1717067956924438,
+      "step": 374
+    },
+    {
+      "epoch": 1.49800796812749,
+      "grad_norm": 0.5834475159645081,
+      "learning_rate": 7.66815151812307e-06,
+      "loss": 1.1137268543243408,
+      "step": 376
+    },
+    {
+      "epoch": 1.5059760956175299,
+      "grad_norm": 0.5452304482460022,
+      "learning_rate": 7.642106588249651e-06,
+      "loss": 0.9259577393531799,
+      "step": 378
+    },
+    {
+      "epoch": 1.5139442231075697,
+      "grad_norm": 0.26578736305236816,
+      "learning_rate": 7.6159685447073376e-06,
+      "loss": 1.12434983253479,
+      "step": 380
+    },
+    {
+      "epoch": 1.5219123505976095,
+      "grad_norm": 0.4347018301486969,
+      "learning_rate": 7.589738523671685e-06,
+      "loss": 0.7884114980697632,
+      "step": 382
+    },
+    {
+      "epoch": 1.5298804780876494,
+      "grad_norm": 0.20773634314537048,
+      "learning_rate": 7.563417665316338e-06,
+      "loss": 0.7225393056869507,
+      "step": 384
+    },
+    {
+      "epoch": 1.5378486055776892,
+      "grad_norm": 0.6731134653091431,
+      "learning_rate": 7.5370071137634984e-06,
+      "loss": 1.0410466194152832,
+      "step": 386
+    },
+    {
+      "epoch": 1.545816733067729,
+      "grad_norm": 0.27709197998046875,
+      "learning_rate": 7.510508017034165e-06,
+      "loss": 1.1276218891143799,
+      "step": 388
+    },
+    {
+      "epoch": 1.5537848605577689,
+      "grad_norm": 0.4856168031692505,
+      "learning_rate": 7.483921526998247e-06,
+      "loss": 0.5739715099334717,
+      "step": 390
+    },
+    {
+      "epoch": 1.5617529880478087,
+      "grad_norm": 0.3228149712085724,
+      "learning_rate": 7.4572487993244865e-06,
+      "loss": 0.5827398300170898,
+      "step": 392
+    },
+    {
+      "epoch": 1.5697211155378485,
+      "grad_norm": 1.0661041736602783,
+      "learning_rate": 7.430490993430232e-06,
+      "loss": 1.1464160680770874,
+      "step": 394
+    },
+    {
+      "epoch": 1.5776892430278884,
+      "grad_norm": 0.47056064009666443,
+      "learning_rate": 7.403649272431027e-06,
+      "loss": 1.0253816843032837,
+      "step": 396
+    },
+    {
+      "epoch": 1.5856573705179282,
+      "grad_norm": 0.2127651870250702,
+      "learning_rate": 7.376724803090069e-06,
+      "loss": 1.1247596740722656,
+      "step": 398
+    },
+    {
+      "epoch": 1.593625498007968,
+      "grad_norm": 0.48643815517425537,
+      "learning_rate": 7.349718755767473e-06,
+      "loss": 0.6272667646408081,
+      "step": 400
+    },
+    {
+      "epoch": 1.6015936254980079,
+      "grad_norm": 0.26240265369415283,
+      "learning_rate": 7.322632304369416e-06,
+      "loss": 1.1022975444793701,
+      "step": 402
+    },
+    {
+      "epoch": 1.6095617529880477,
+      "grad_norm": 0.2799125611782074,
+      "learning_rate": 7.2954666262970965e-06,
+      "loss": 1.248570442199707,
+      "step": 404
+    },
+    {
+      "epoch": 1.6175298804780875,
+      "grad_norm": 0.41543102264404297,
+      "learning_rate": 7.268222902395562e-06,
+      "loss": 0.7265171408653259,
+      "step": 406
+    },
+    {
+      "epoch": 1.6254980079681274,
+      "grad_norm": 0.278363972902298,
+      "learning_rate": 7.240902316902373e-06,
+      "loss": 0.5697952508926392,
+      "step": 408
+    },
+    {
+      "epoch": 1.6334661354581672,
+      "grad_norm": 0.24566787481307983,
+      "learning_rate": 7.213506057396134e-06,
+      "loss": 0.752501368522644,
+      "step": 410
+    },
+    {
+      "epoch": 1.641434262948207,
+      "grad_norm": 0.6162587404251099,
+      "learning_rate": 7.186035314744872e-06,
+      "loss": 0.5186296701431274,
+      "step": 412
+    },
+    {
+      "epoch": 1.6494023904382469,
+      "grad_norm": 0.26647666096687317,
+      "learning_rate": 7.1584912830542575e-06,
+      "loss": 0.80875164270401,
+      "step": 414
+    },
+    {
+      "epoch": 1.6573705179282867,
+      "grad_norm": 0.34343570470809937,
+      "learning_rate": 7.130875159615716e-06,
+      "loss": 0.7356295585632324,
+      "step": 416
+    },
+    {
+      "epoch": 1.6653386454183265,
+      "grad_norm": 0.4066859781742096,
+      "learning_rate": 7.103188144854376e-06,
+      "loss": 0.938122034072876,
+      "step": 418
+    },
+    {
+      "epoch": 1.6733067729083664,
+      "grad_norm": 0.3127875328063965,
+      "learning_rate": 7.075431442276888e-06,
+      "loss": 1.1433111429214478,
+      "step": 420
+    },
+    {
+      "epoch": 1.6812749003984062,
+      "grad_norm": 0.2553024888038635,
+      "learning_rate": 7.047606258419108e-06,
+      "loss": 1.0093721151351929,
+      "step": 422
+    },
+    {
+      "epoch": 1.6892430278884463,
+      "grad_norm": 1.2878453731536865,
+      "learning_rate": 7.019713802793663e-06,
+      "loss": 0.5690187215805054,
+      "step": 424
+    },
+    {
+      "epoch": 1.697211155378486,
+      "grad_norm": 0.2481621950864792,
+      "learning_rate": 6.99175528783736e-06,
+      "loss": 1.1795101165771484,
+      "step": 426
+    },
+    {
+      "epoch": 1.705179282868526,
+      "grad_norm": 0.31707319617271423,
+      "learning_rate": 6.963731928858495e-06,
+      "loss": 1.1185859441757202,
+      "step": 428
+    },
+    {
+      "epoch": 1.7131474103585658,
+      "grad_norm": 0.3519105017185211,
+      "learning_rate": 6.935644943984017e-06,
+      "loss": 0.7240161895751953,
+      "step": 430
+    },
+    {
+      "epoch": 1.7211155378486056,
+      "grad_norm": 0.62026047706604,
+      "learning_rate": 6.9074955541065865e-06,
+      "loss": 0.873826265335083,
+      "step": 432
+    },
+    {
+      "epoch": 1.7290836653386454,
+      "grad_norm": 0.8380200862884521,
+      "learning_rate": 6.8792849828315054e-06,
+      "loss": 0.6440986394882202,
+      "step": 434
+    },
+    {
+      "epoch": 1.7370517928286853,
+      "grad_norm": 1.0566929578781128,
+      "learning_rate": 6.851014456423515e-06,
+      "loss": 0.3889099359512329,
+      "step": 436
+    },
+    {
+      "epoch": 1.745019920318725,
+      "grad_norm": 2.003751754760742,
+      "learning_rate": 6.822685203753508e-06,
+      "loss": 0.7689342498779297,
+      "step": 438
+    },
+    {
+      "epoch": 1.752988047808765,
+      "grad_norm": 1.1321049928665161,
+      "learning_rate": 6.794298456245111e-06,
+      "loss": 0.858902096748352,
+      "step": 440
+    },
+    {
+      "epoch": 1.7609561752988048,
+      "grad_norm": 0.5031111836433411,
+      "learning_rate": 6.765855447821144e-06,
+      "loss": 1.4214740991592407,
+      "step": 442
+    },
+    {
+      "epoch": 1.7689243027888446,
+      "grad_norm": 0.27599719166755676,
+      "learning_rate": 6.737357414849993e-06,
+      "loss": 0.7524920105934143,
+      "step": 444
+    },
+    {
+      "epoch": 1.7768924302788844,
+      "grad_norm": 0.2588093876838684,
+      "learning_rate": 6.70880559609187e-06,
+      "loss": 1.0499422550201416,
+      "step": 446
+    },
+    {
+      "epoch": 1.7848605577689243,
+      "grad_norm": 0.2442282885313034,
+      "learning_rate": 6.680201232644958e-06,
+      "loss": 1.1038804054260254,
+      "step": 448
+    },
+    {
+      "epoch": 1.792828685258964,
+      "grad_norm": 0.21524015069007874,
+      "learning_rate": 6.651545567891471e-06,
+      "loss": 0.8592348098754883,
+      "step": 450
+    },
+    {
+      "epoch": 1.800796812749004,
+      "grad_norm": 0.15642908215522766,
+      "learning_rate": 6.622839847443596e-06,
+      "loss": 0.47004958987236023,
+      "step": 452
+    },
+    {
+      "epoch": 1.8087649402390438,
+      "grad_norm": 0.4115596115589142,
+      "learning_rate": 6.59408531908936e-06,
+      "loss": 0.9372057318687439,
+      "step": 454
+    },
+    {
+      "epoch": 1.8167330677290838,
+      "grad_norm": 1.0596694946289062,
+      "learning_rate": 6.565283232738386e-06,
+      "loss": 0.7479291558265686,
+      "step": 456
+    },
+    {
+      "epoch": 1.8247011952191237,
+      "grad_norm": 0.34829071164131165,
+      "learning_rate": 6.536434840367555e-06,
+      "loss": 1.0391331911087036,
+      "step": 458
+    },
+    {
+      "epoch": 1.8326693227091635,
+      "grad_norm": 0.42441603541374207,
+      "learning_rate": 6.507541395966591e-06,
+      "loss": 0.7797434329986572,
+      "step": 460
+    },
+    {
+      "epoch": 1.8406374501992033,
+      "grad_norm": 0.20400914549827576,
+      "learning_rate": 6.47860415548356e-06,
+      "loss": 0.8128412961959839,
+      "step": 462
+    },
+    {
+      "epoch": 1.8486055776892432,
+      "grad_norm": 0.16058732569217682,
+      "learning_rate": 6.449624376770259e-06,
+      "loss": 0.12251711636781693,
+      "step": 464
+    },
+    {
+      "epoch": 1.856573705179283,
+      "grad_norm": 0.2379097193479538,
+      "learning_rate": 6.420603319527554e-06,
+      "loss": 1.1542426347732544,
+      "step": 466
+    },
+    {
+      "epoch": 1.8645418326693228,
+      "grad_norm": 1.1358978748321533,
+      "learning_rate": 6.3915422452506146e-06,
+      "loss": 0.5193546414375305,
+      "step": 468
+    },
+    {
+      "epoch": 1.8725099601593627,
+      "grad_norm": 0.24465781450271606,
+      "learning_rate": 6.362442417174082e-06,
+      "loss": 1.1340575218200684,
+      "step": 470
+    },
+    {
+      "epoch": 1.8804780876494025,
+      "grad_norm": 0.7936683297157288,
+      "learning_rate": 6.33330510021716e-06,
+      "loss": 1.059372901916504,
+      "step": 472
+    },
+    {
+      "epoch": 1.8884462151394423,
+      "grad_norm": 9.454405784606934,
+      "learning_rate": 6.304131560928622e-06,
+      "loss": 1.104183554649353,
+      "step": 474
+    },
+    {
+      "epoch": 1.8964143426294822,
+      "grad_norm": 0.21524909138679504,
+      "learning_rate": 6.274923067431775e-06,
+      "loss": 0.9676142930984497,
+      "step": 476
+    },
+    {
+      "epoch": 1.904382470119522,
+      "grad_norm": 0.22511586546897888,
+      "learning_rate": 6.2456808893693155e-06,
+      "loss": 0.83753901720047,
+      "step": 478
+    },
+    {
+      "epoch": 1.9123505976095618,
+      "grad_norm": 0.9536860585212708,
+      "learning_rate": 6.216406297848155e-06,
+      "loss": 0.5888479351997375,
+      "step": 480
+    },
+    {
+      "epoch": 1.9203187250996017,
+      "grad_norm": 0.3726925551891327,
+      "learning_rate": 6.187100565384162e-06,
+      "loss": 0.7634333968162537,
+      "step": 482
+    },
+    {
+      "epoch": 1.9282868525896415,
+      "grad_norm": 0.26928794384002686,
+      "learning_rate": 6.157764965846841e-06,
+      "loss": 1.099297046661377,
+      "step": 484
+    },
+    {
+      "epoch": 1.9362549800796813,
+      "grad_norm": 0.2715873122215271,
+      "learning_rate": 6.128400774403974e-06,
+      "loss": 1.1109613180160522,
+      "step": 486
+    },
+    {
+      "epoch": 1.9442231075697212,
+      "grad_norm": 0.19929516315460205,
+      "learning_rate": 6.0990092674661834e-06,
+      "loss": 1.0884331464767456,
+      "step": 488
+    },
+    {
+      "epoch": 1.952191235059761,
+      "grad_norm": 0.1862286478281021,
+      "learning_rate": 6.069591722631446e-06,
+      "loss": 1.0938901901245117,
+      "step": 490
+    },
+    {
+      "epoch": 1.9601593625498008,
+      "grad_norm": 0.437503844499588,
+      "learning_rate": 6.040149418629562e-06,
+      "loss": 0.7570807933807373,
+      "step": 492
+    },
+    {
+      "epoch": 1.9681274900398407,
+      "grad_norm": 0.8427123427391052,
+      "learning_rate": 6.010683635266569e-06,
+      "loss": 0.7201614379882812,
+      "step": 494
+    },
+    {
+      "epoch": 1.9760956175298805,
+      "grad_norm": 0.2153085321187973,
+      "learning_rate": 5.981195653369113e-06,
+      "loss": 1.061843991279602,
+      "step": 496
+    },
+    {
+      "epoch": 1.9840637450199203,
+      "grad_norm": 2.7114696502685547,
+      "learning_rate": 5.95168675472877e-06,
+      "loss": 0.8539069294929504,
+      "step": 498
+    },
+    {
+      "epoch": 1.9920318725099602,
+      "grad_norm": 0.2182704508304596,
+      "learning_rate": 5.922158222046334e-06,
+      "loss": 1.1612799167633057,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3179067075252533,
+      "learning_rate": 5.892611338876053e-06,
+      "loss": 0.6118001341819763,
+      "step": 502
+    },
+    {
+      "epoch": 2.00796812749004,
+      "grad_norm": 0.6945406794548035,
+      "learning_rate": 5.8630473895698426e-06,
+      "loss": 0.5886072516441345,
+      "step": 504
+    },
+    {
+      "epoch": 2.0159362549800797,
+      "grad_norm": 0.5343687534332275,
+      "learning_rate": 5.833467659221452e-06,
+      "loss": 0.503696620464325,
+      "step": 506
+    },
+    {
+      "epoch": 2.0239043824701195,
+      "grad_norm": 0.0851997509598732,
+      "learning_rate": 5.803873433610604e-06,
+      "loss": 0.5175277590751648,
+      "step": 508
+    },
+    {
+      "epoch": 2.0318725099601593,
+      "grad_norm": 0.20901891589164734,
+      "learning_rate": 5.774265999147108e-06,
+      "loss": 0.8128920197486877,
+      "step": 510
+    },
+    {
+      "epoch": 2.039840637450199,
+      "grad_norm": 0.5096555352210999,
+      "learning_rate": 5.744646642814935e-06,
+      "loss": 0.696502685546875,
+      "step": 512
+    },
+    {
+      "epoch": 2.047808764940239,
+      "grad_norm": 2.1485402584075928,
+      "learning_rate": 5.715016652116284e-06,
+      "loss": 0.3971967399120331,
+      "step": 514
+    },
+    {
+      "epoch": 2.055776892430279,
+      "grad_norm": 0.22098702192306519,
+      "learning_rate": 5.68537731501561e-06,
+      "loss": 0.8687353730201721,
+      "step": 516
+    },
+    {
+      "epoch": 2.0637450199203187,
+      "grad_norm": 0.7635393142700195,
+      "learning_rate": 5.655729919883639e-06,
+      "loss": 0.5329861044883728,
+      "step": 518
+    },
+    {
+      "epoch": 2.0717131474103585,
+      "grad_norm": 0.5440754890441895,
+      "learning_rate": 5.626075755441366e-06,
+      "loss": 0.7716398239135742,
+      "step": 520
+    },
+    {
+      "epoch": 2.0796812749003983,
+      "grad_norm": 0.3139892518520355,
+      "learning_rate": 5.596416110704034e-06,
+      "loss": 0.9331838488578796,
+      "step": 522
+    },
+    {
+      "epoch": 2.087649402390438,
+      "grad_norm": 0.6210455894470215,
+      "learning_rate": 5.566752274925109e-06,
+      "loss": 0.8374670147895813,
+      "step": 524
+    },
+    {
+      "epoch": 2.095617529880478,
+      "grad_norm": 0.852881133556366,
+      "learning_rate": 5.5370855375402345e-06,
+      "loss": 0.4146691560745239,
+      "step": 526
+    },
+    {
+      "epoch": 2.103585657370518,
+      "grad_norm": 0.2397157996892929,
+      "learning_rate": 5.507417188111176e-06,
+      "loss": 0.7079000473022461,
+      "step": 528
+    },
+    {
+      "epoch": 2.1115537848605577,
+      "grad_norm": 0.16977767646312714,
+      "learning_rate": 5.477748516269779e-06,
+      "loss": 0.5761277079582214,
+      "step": 530
+    },
+    {
+      "epoch": 2.1195219123505975,
+      "grad_norm": 0.8316014409065247,
+      "learning_rate": 5.448080811661897e-06,
+      "loss": 0.22371451556682587,
+      "step": 532
+    },
+    {
+      "epoch": 2.1274900398406373,
+      "grad_norm": 0.2960129380226135,
+      "learning_rate": 5.418415363891346e-06,
+      "loss": 0.7184697985649109,
+      "step": 534
+    },
+    {
+      "epoch": 2.135458167330677,
+      "grad_norm": 0.6863223910331726,
+      "learning_rate": 5.3887534624638335e-06,
+      "loss": 0.6520018577575684,
+      "step": 536
+    },
+    {
+      "epoch": 2.143426294820717,
+      "grad_norm": 1.7351683378219604,
+      "learning_rate": 5.359096396730924e-06,
+      "loss": 0.684842586517334,
+      "step": 538
+    },
+    {
+      "epoch": 2.151394422310757,
+      "grad_norm": 0.21653281152248383,
+      "learning_rate": 5.329445455833975e-06,
+      "loss": 0.9097157120704651,
+      "step": 540
+    },
+    {
+      "epoch": 2.1593625498007967,
+      "grad_norm": 0.26099443435668945,
+      "learning_rate": 5.299801928648105e-06,
+      "loss": 0.8514461517333984,
+      "step": 542
+    },
+    {
+      "epoch": 2.1673306772908365,
+      "grad_norm": 0.30788734555244446,
+      "learning_rate": 5.270167103726181e-06,
+      "loss": 0.681281328201294,
+      "step": 544
+    },
+    {
+      "epoch": 2.1752988047808763,
+      "grad_norm": 0.24177972972393036,
+      "learning_rate": 5.240542269242794e-06,
+      "loss": 0.6330439448356628,
+      "step": 546
+    },
+    {
+      "epoch": 2.183266932270916,
+      "grad_norm": 0.2683347165584564,
+      "learning_rate": 5.210928712938262e-06,
+      "loss": 0.9867715835571289,
+      "step": 548
+    },
+    {
+      "epoch": 2.191235059760956,
+      "grad_norm": 0.21875284612178802,
+      "learning_rate": 5.181327722062667e-06,
+      "loss": 0.9723871350288391,
+      "step": 550
+    },
+    {
+      "epoch": 2.199203187250996,
+      "grad_norm": 0.4430513083934784,
+      "learning_rate": 5.151740583319892e-06,
+      "loss": 0.6143315434455872,
+      "step": 552
+    },
+    {
+      "epoch": 2.2071713147410357,
+      "grad_norm": 0.679879903793335,
+      "learning_rate": 5.122168582811689e-06,
+      "loss": 0.4490898847579956,
+      "step": 554
+    },
+    {
+      "epoch": 2.2151394422310755,
+      "grad_norm": 0.2948717772960663,
+      "learning_rate": 5.092613005981781e-06,
+      "loss": 0.6454632878303528,
+      "step": 556
+    },
+    {
+      "epoch": 2.2231075697211153,
+      "grad_norm": 0.4709537625312805,
+      "learning_rate": 5.063075137559979e-06,
+      "loss": 0.8263673782348633,
+      "step": 558
+    },
+    {
+      "epoch": 2.231075697211155,
+      "grad_norm": 0.23535384237766266,
+      "learning_rate": 5.033556261506343e-06,
+      "loss": 0.8837491273880005,
+      "step": 560
+    },
+    {
+      "epoch": 2.239043824701195,
+      "grad_norm": 0.2577301561832428,
+      "learning_rate": 5.0040576609553645e-06,
+      "loss": 0.9912109971046448,
+      "step": 562
+    },
+    {
+      "epoch": 2.247011952191235,
+      "grad_norm": 0.4326441287994385,
+      "learning_rate": 4.974580618160194e-06,
+      "loss": 0.7440586090087891,
+      "step": 564
+    },
+    {
+      "epoch": 2.2549800796812747,
+      "grad_norm": 0.7311254739761353,
+      "learning_rate": 4.945126414436908e-06,
+      "loss": 0.5915436148643494,
+      "step": 566
+    },
+    {
+      "epoch": 2.2629482071713145,
+      "grad_norm": 0.192102313041687,
+      "learning_rate": 4.915696330108804e-06,
+      "loss": 0.7124112844467163,
+      "step": 568
+    },
+    {
+      "epoch": 2.2709163346613543,
+      "grad_norm": 0.2934555113315582,
+      "learning_rate": 4.886291644450751e-06,
+      "loss": 0.943176805973053,
+      "step": 570
+    },
+    {
+      "epoch": 2.278884462151394,
+      "grad_norm": 0.5654205083847046,
+      "learning_rate": 4.8569136356335845e-06,
+      "loss": 0.6194332242012024,
+      "step": 572
+    },
+    {
+      "epoch": 2.2868525896414345,
+      "grad_norm": 0.36352983117103577,
+      "learning_rate": 4.827563580668544e-06,
+      "loss": 0.6795768141746521,
+      "step": 574
+    },
+    {
+      "epoch": 2.2948207171314743,
+      "grad_norm": 0.21166585385799408,
+      "learning_rate": 4.798242755351755e-06,
+      "loss": 0.23236216604709625,
+      "step": 576
+    },
+    {
+      "epoch": 2.302788844621514,
+      "grad_norm": 0.3600130081176758,
+      "learning_rate": 4.768952434208795e-06,
+      "loss": 0.6390988230705261,
+      "step": 578
+    },
+    {
+      "epoch": 2.310756972111554,
+      "grad_norm": 0.28703030943870544,
+      "learning_rate": 4.7396938904392656e-06,
+      "loss": 0.6285381317138672,
+      "step": 580
+    },
+    {
+      "epoch": 2.318725099601594,
+      "grad_norm": 0.5645820498466492,
+      "learning_rate": 4.71046839586147e-06,
+      "loss": 0.5717098116874695,
+      "step": 582
+    },
+    {
+      "epoch": 2.3266932270916336,
+      "grad_norm": 0.2184012234210968,
+      "learning_rate": 4.681277220857108e-06,
+      "loss": 0.7276479005813599,
+      "step": 584
+    },
+    {
+      "epoch": 2.3346613545816735,
+      "grad_norm": 0.22574515640735626,
+      "learning_rate": 4.6521216343160824e-06,
+      "loss": 0.6201709508895874,
+      "step": 586
+    },
+    {
+      "epoch": 2.3426294820717133,
+      "grad_norm": 0.28968313336372375,
+      "learning_rate": 4.623002903581316e-06,
+      "loss": 0.8008480668067932,
+      "step": 588
+    },
+    {
+      "epoch": 2.350597609561753,
+      "grad_norm": 0.24634480476379395,
+      "learning_rate": 4.593922294393677e-06,
+      "loss": 0.9472393989562988,
+      "step": 590
+    },
+    {
+      "epoch": 2.358565737051793,
+      "grad_norm": 0.2618328034877777,
+      "learning_rate": 4.564881070836955e-06,
+      "loss": 0.9737678170204163,
+      "step": 592
+    },
+    {
+      "epoch": 2.366533864541833,
+      "grad_norm": 0.3499445915222168,
+      "learning_rate": 4.535880495282916e-06,
+      "loss": 0.6368265151977539,
+      "step": 594
+    },
+    {
+      "epoch": 2.3745019920318726,
+      "grad_norm": 0.2279530018568039,
+      "learning_rate": 4.506921828336427e-06,
+      "loss": 0.9475688934326172,
+      "step": 596
+    },
+    {
+      "epoch": 2.3824701195219125,
+      "grad_norm": 0.2167678326368332,
+      "learning_rate": 4.478006328780657e-06,
+      "loss": 0.6551421284675598,
+      "step": 598
+    },
+    {
+      "epoch": 2.3904382470119523,
+      "grad_norm": 1.3437480926513672,
+      "learning_rate": 4.44913525352237e-06,
+      "loss": 0.8309920430183411,
+      "step": 600
+    },
+    {
+      "epoch": 2.398406374501992,
+      "grad_norm": 0.23248842358589172,
+      "learning_rate": 4.420309857537277e-06,
+      "loss": 0.9577336311340332,
+      "step": 602
+    },
+    {
+      "epoch": 2.406374501992032,
+      "grad_norm": 0.2530728578567505,
+      "learning_rate": 4.39153139381549e-06,
+      "loss": 0.5325211882591248,
+      "step": 604
+    },
+    {
+      "epoch": 2.414342629482072,
+      "grad_norm": 0.5845907926559448,
+      "learning_rate": 4.3628011133070605e-06,
+      "loss": 0.2235029935836792,
+      "step": 606
+    },
+    {
+      "epoch": 2.4223107569721116,
+      "grad_norm": 0.2823912799358368,
+      "learning_rate": 4.334120264867596e-06,
+      "loss": 0.45812419056892395,
+      "step": 608
+    },
+    {
+      "epoch": 2.4302788844621515,
+      "grad_norm": 0.15544608235359192,
+      "learning_rate": 4.305490095203979e-06,
+      "loss": 0.8220903873443604,
+      "step": 610
+    },
+    {
+      "epoch": 2.4382470119521913,
+      "grad_norm": 0.24993936717510223,
+      "learning_rate": 4.27691184882017e-06,
+      "loss": 0.7239663004875183,
+      "step": 612
+    },
+    {
+      "epoch": 2.446215139442231,
+      "grad_norm": 0.2656114399433136,
+      "learning_rate": 4.248386767963121e-06,
+      "loss": 0.5796704292297363,
+      "step": 614
+    },
+    {
+      "epoch": 2.454183266932271,
+      "grad_norm": 0.23990172147750854,
+      "learning_rate": 4.219916092568764e-06,
+      "loss": 0.836432695388794,
+      "step": 616
+    },
+    {
+      "epoch": 2.462151394422311,
+      "grad_norm": 0.5075326561927795,
+      "learning_rate": 4.191501060208121e-06,
+      "loss": 0.6542029976844788,
+      "step": 618
+    },
+    {
+      "epoch": 2.4701195219123506,
+      "grad_norm": 0.42840975522994995,
+      "learning_rate": 4.163142906033512e-06,
+      "loss": 0.7549333572387695,
+      "step": 620
+    },
+    {
+      "epoch": 2.4780876494023905,
+      "grad_norm": 0.26510748267173767,
+      "learning_rate": 4.134842862724854e-06,
+      "loss": 0.7886976003646851,
+      "step": 622
+    },
+    {
+      "epoch": 2.4860557768924303,
+      "grad_norm": 0.3405495584011078,
+      "learning_rate": 4.106602160436088e-06,
+      "loss": 0.735430896282196,
+      "step": 624
+    },
+    {
+      "epoch": 2.49402390438247,
+      "grad_norm": 0.3593902289867401,
+      "learning_rate": 4.078422026741706e-06,
+      "loss": 0.5373623967170715,
+      "step": 626
+    },
+    {
+      "epoch": 2.50199203187251,
+      "grad_norm": 0.4967736303806305,
+      "learning_rate": 4.050303686583386e-06,
+      "loss": 0.938071608543396,
+      "step": 628
+    },
+    {
+      "epoch": 2.50996015936255,
+      "grad_norm": 0.40400850772857666,
+      "learning_rate": 4.0222483622167465e-06,
+      "loss": 0.42294827103614807,
+      "step": 630
+    },
+    {
+      "epoch": 2.5179282868525896,
+      "grad_norm": 3.144519329071045,
+      "learning_rate": 3.9942572731582185e-06,
+      "loss": 0.4069959223270416,
+      "step": 632
+    },
+    {
+      "epoch": 2.5258964143426295,
+      "grad_norm": 0.26504188776016235,
+      "learning_rate": 3.966331636132039e-06,
+      "loss": 0.9584428071975708,
+      "step": 634
+    },
+    {
+      "epoch": 2.5338645418326693,
+      "grad_norm": 0.33528634905815125,
+      "learning_rate": 3.938472665017355e-06,
+      "loss": 0.8962292671203613,
+      "step": 636
+    },
+    {
+      "epoch": 2.541832669322709,
+      "grad_norm": 0.5937342643737793,
+      "learning_rate": 3.91068157079546e-06,
+      "loss": 0.9537577629089355,
+      "step": 638
+    },
+    {
+      "epoch": 2.549800796812749,
+      "grad_norm": 0.1715266853570938,
+      "learning_rate": 3.882959561497159e-06,
+      "loss": 0.4940541088581085,
+      "step": 640
+    },
+    {
+      "epoch": 2.557768924302789,
+      "grad_norm": 0.6008433699607849,
+      "learning_rate": 3.855307842150259e-06,
+      "loss": 0.907759964466095,
+      "step": 642
+    },
+    {
+      "epoch": 2.5657370517928286,
+      "grad_norm": 0.34849610924720764,
+      "learning_rate": 3.827727614727177e-06,
+      "loss": 0.9829540252685547,
+      "step": 644
+    },
+    {
+      "epoch": 2.5737051792828685,
+      "grad_norm": 0.2508140206336975,
+      "learning_rate": 3.8002200780927e-06,
+      "loss": 0.6158788204193115,
+      "step": 646
+    },
+    {
+      "epoch": 2.5816733067729083,
+      "grad_norm": 0.9183971881866455,
+      "learning_rate": 3.772786427951878e-06,
+      "loss": 0.9395154714584351,
+      "step": 648
+    },
+    {
+      "epoch": 2.589641434262948,
+      "grad_norm": 0.34092074632644653,
+      "learning_rate": 3.7454278567980406e-06,
+      "loss": 0.7144474983215332,
+      "step": 650
+    },
+    {
+      "epoch": 2.597609561752988,
+      "grad_norm": 0.8458039164543152,
+      "learning_rate": 3.7181455538609577e-06,
+      "loss": 0.7298926711082458,
+      "step": 652
+    },
+    {
+      "epoch": 2.605577689243028,
+      "grad_norm": 0.9377502799034119,
+      "learning_rate": 3.690940705055163e-06,
+      "loss": 0.6572567224502563,
+      "step": 654
+    },
+    {
+      "epoch": 2.6135458167330676,
+      "grad_norm": 0.4728657603263855,
+      "learning_rate": 3.6638144929283824e-06,
+      "loss": 0.32404664158821106,
+      "step": 656
+    },
+    {
+      "epoch": 2.6215139442231075,
+      "grad_norm": 0.3605116009712219,
+      "learning_rate": 3.6367680966101483e-06,
+      "loss": 0.5177608728408813,
+      "step": 658
+    },
+    {
+      "epoch": 2.6294820717131473,
+      "grad_norm": 0.264681875705719,
+      "learning_rate": 3.6098026917605334e-06,
+      "loss": 0.8240447044372559,
+      "step": 660
+    },
+    {
+      "epoch": 2.637450199203187,
+      "grad_norm": 1.118610143661499,
+      "learning_rate": 3.5829194505190534e-06,
+      "loss": 0.8320629596710205,
+      "step": 662
+    },
+    {
+      "epoch": 2.645418326693227,
+      "grad_norm": 0.4917624294757843,
+      "learning_rate": 3.5561195414537164e-06,
+      "loss": 0.5248578786849976,
+      "step": 664
+    },
+    {
+      "epoch": 2.653386454183267,
+      "grad_norm": 0.3855407238006592,
+      "learning_rate": 3.529404129510221e-06,
+      "loss": 0.8523591756820679,
+      "step": 666
+    },
+    {
+      "epoch": 2.6613545816733066,
+      "grad_norm": 0.1237802505493164,
+      "learning_rate": 3.502774375961321e-06,
+      "loss": 0.585418164730072,
+      "step": 668
+    },
+    {
+      "epoch": 2.6693227091633465,
+      "grad_norm": 0.35880449414253235,
+      "learning_rate": 3.4762314383563533e-06,
+      "loss": 0.7037955522537231,
+      "step": 670
+    },
+    {
+      "epoch": 2.6772908366533863,
+      "grad_norm": 0.122527576982975,
+      "learning_rate": 3.4497764704709143e-06,
+      "loss": 0.33243533968925476,
+      "step": 672
+    },
+    {
+      "epoch": 2.685258964143426,
+      "grad_norm": 0.22298847138881683,
+      "learning_rate": 3.4234106222567055e-06,
+      "loss": 0.9578478336334229,
+      "step": 674
+    },
+    {
+      "epoch": 2.6932270916334664,
+      "grad_norm": 0.47459158301353455,
+      "learning_rate": 3.3971350397915504e-06,
+      "loss": 0.6586756706237793,
+      "step": 676
+    },
+    {
+      "epoch": 2.7011952191235062,
+      "grad_norm": 0.3859528601169586,
+      "learning_rate": 3.370950865229583e-06,
+      "loss": 0.5740290284156799,
+      "step": 678
+    },
+    {
+      "epoch": 2.709163346613546,
+      "grad_norm": 0.17183351516723633,
+      "learning_rate": 3.344859236751585e-06,
+      "loss": 0.9585554599761963,
+      "step": 680
+    },
+    {
+      "epoch": 2.717131474103586,
+      "grad_norm": 0.18819180130958557,
+      "learning_rate": 3.3188612885155257e-06,
+      "loss": 0.981060802936554,
+      "step": 682
+    },
+    {
+      "epoch": 2.7250996015936257,
+      "grad_norm": 0.37404805421829224,
+      "learning_rate": 3.2929581506072596e-06,
+      "loss": 0.6816679835319519,
+      "step": 684
+    },
+    {
+      "epoch": 2.7330677290836656,
+      "grad_norm": 0.734767735004425,
+      "learning_rate": 3.267150948991387e-06,
+      "loss": 0.44262567162513733,
+      "step": 686
+    },
+    {
+      "epoch": 2.7410358565737054,
+      "grad_norm": 0.23475851118564606,
+      "learning_rate": 3.241440805462337e-06,
+      "loss": 1.0185444355010986,
+      "step": 688
+    },
+    {
+      "epoch": 2.7490039840637452,
+      "grad_norm": 1.8315621614456177,
+      "learning_rate": 3.215828837595585e-06,
+      "loss": 0.49544572830200195,
+      "step": 690
+    },
+    {
+      "epoch": 2.756972111553785,
+      "grad_norm": 0.2423868477344513,
+      "learning_rate": 3.190316158699084e-06,
+      "loss": 1.045667052268982,
+      "step": 692
+    },
+    {
+      "epoch": 2.764940239043825,
+      "grad_norm": 0.31695112586021423,
+      "learning_rate": 3.1649038777648622e-06,
+      "loss": 0.8049361109733582,
+      "step": 694
+    },
+    {
+      "epoch": 2.7729083665338647,
+      "grad_norm": 0.6442317366600037,
+      "learning_rate": 3.1395930994208234e-06,
+      "loss": 0.6994091868400574,
+      "step": 696
+    },
+    {
+      "epoch": 2.7808764940239046,
+      "grad_norm": 0.2618781626224518,
+      "learning_rate": 3.114384923882734e-06,
+      "loss": 0.8016667366027832,
+      "step": 698
+    },
+    {
+      "epoch": 2.7888446215139444,
+      "grad_norm": 0.2563982605934143,
+      "learning_rate": 3.089280446906387e-06,
+      "loss": 0.9851805567741394,
+      "step": 700
+    },
+    {
+      "epoch": 2.7968127490039842,
+      "grad_norm": 0.3183526396751404,
+      "learning_rate": 3.0642807597399847e-06,
+      "loss": 0.5757198929786682,
+      "step": 702
+    },
+    {
+      "epoch": 2.804780876494024,
+      "grad_norm": 0.443581223487854,
+      "learning_rate": 3.0393869490766913e-06,
+      "loss": 1.1037681102752686,
+      "step": 704
+    },
+    {
+      "epoch": 2.812749003984064,
+      "grad_norm": 0.16616466641426086,
+      "learning_rate": 3.0146000970074087e-06,
+      "loss": 1.0674716234207153,
+      "step": 706
+    },
+    {
+      "epoch": 2.8207171314741037,
+      "grad_norm": 0.33500173687934875,
+      "learning_rate": 2.9899212809737275e-06,
+      "loss": 1.0235587358474731,
+      "step": 708
+    },
+    {
+      "epoch": 2.8286852589641436,
+      "grad_norm": 0.2945948541164398,
+      "learning_rate": 2.9653515737211037e-06,
+      "loss": 0.5113028883934021,
+      "step": 710
+    },
+    {
+      "epoch": 2.8366533864541834,
+      "grad_norm": 0.23376336693763733,
+      "learning_rate": 2.940892043252225e-06,
+      "loss": 1.0040061473846436,
+      "step": 712
+    },
+    {
+      "epoch": 2.8446215139442232,
+      "grad_norm": 0.20928576588630676,
+      "learning_rate": 2.916543752780574e-06,
+      "loss": 0.5935018062591553,
+      "step": 714
+    },
+    {
+      "epoch": 2.852589641434263,
+      "grad_norm": 0.23121950030326843,
+      "learning_rate": 2.892307760684234e-06,
+      "loss": 0.9517689347267151,
+      "step": 716
+    },
+    {
+      "epoch": 2.860557768924303,
+      "grad_norm": 0.5248010754585266,
+      "learning_rate": 2.8681851204598675e-06,
+      "loss": 0.5970214009284973,
+      "step": 718
+    },
+    {
+      "epoch": 2.8685258964143427,
+      "grad_norm": 0.2068823128938675,
+      "learning_rate": 2.844176880676931e-06,
+      "loss": 0.5916792750358582,
+      "step": 720
+    },
+    {
+      "epoch": 2.8764940239043826,
+      "grad_norm": 0.32244107127189636,
+      "learning_rate": 2.8202840849320867e-06,
+      "loss": 0.6319440603256226,
+      "step": 722
+    },
+    {
+      "epoch": 2.8844621513944224,
+      "grad_norm": 0.249526709318161,
+      "learning_rate": 2.7965077718038437e-06,
+      "loss": 0.6140936613082886,
+      "step": 724
+    },
+    {
+      "epoch": 2.8924302788844622,
+      "grad_norm": 0.2989876866340637,
+      "learning_rate": 2.7728489748074193e-06,
+      "loss": 0.9145041704177856,
+      "step": 726
+    },
+    {
+      "epoch": 2.900398406374502,
+      "grad_norm": 0.295612096786499,
+      "learning_rate": 2.7493087223498015e-06,
+      "loss": 0.5431255102157593,
+      "step": 728
+    },
+    {
+      "epoch": 2.908366533864542,
+      "grad_norm": 0.24679818749427795,
+      "learning_rate": 2.725888037685057e-06,
+      "loss": 0.9023692607879639,
+      "step": 730
+    },
+    {
+      "epoch": 2.9163346613545817,
+      "grad_norm": 0.27754729986190796,
+      "learning_rate": 2.702587938869842e-06,
+      "loss": 0.3047316372394562,
+      "step": 732
+    },
+    {
+      "epoch": 2.9243027888446216,
+      "grad_norm": 0.29821181297302246,
+      "learning_rate": 2.6794094387191617e-06,
+      "loss": 0.9856127500534058,
+      "step": 734
+    },
+    {
+      "epoch": 2.9322709163346614,
+      "grad_norm": 0.5145055651664734,
+      "learning_rate": 2.6563535447623302e-06,
+      "loss": 0.42886558175086975,
+      "step": 736
+    },
+    {
+      "epoch": 2.9402390438247012,
+      "grad_norm": 0.7824696898460388,
+      "learning_rate": 2.6334212591991847e-06,
+      "loss": 0.8161010146141052,
+      "step": 738
+    },
+    {
+      "epoch": 2.948207171314741,
+      "grad_norm": 0.18964532017707825,
+      "learning_rate": 2.6106135788565248e-06,
+      "loss": 1.0531599521636963,
+      "step": 740
+    },
+    {
+      "epoch": 2.956175298804781,
+      "grad_norm": 1.3559470176696777,
+      "learning_rate": 2.5879314951447688e-06,
+      "loss": 0.3177207410335541,
+      "step": 742
+    },
+    {
+      "epoch": 2.9641434262948207,
+      "grad_norm": 0.20690570771694183,
+      "learning_rate": 2.5653759940148692e-06,
+      "loss": 0.8213226199150085,
+      "step": 744
+    },
+    {
+      "epoch": 2.9721115537848606,
+      "grad_norm": 0.24971678853034973,
+      "learning_rate": 2.542948055915454e-06,
+      "loss": 0.8660821914672852,
+      "step": 746
+    },
+    {
+      "epoch": 2.9800796812749004,
+      "grad_norm": 0.4176087975502014,
+      "learning_rate": 2.5206486557502087e-06,
+      "loss": 0.47795814275741577,
+      "step": 748
+    },
+    {
+      "epoch": 2.9880478087649402,
+      "grad_norm": 0.3064267933368683,
+      "learning_rate": 2.498478762835492e-06,
+      "loss": 0.14798124134540558,
+      "step": 750
+    },
+    {
+      "epoch": 2.99601593625498,
+      "grad_norm": 0.276616632938385,
+      "learning_rate": 2.476439340858208e-06,
+      "loss": 0.9450386762619019,
+      "step": 752
+    },
+    {
+      "epoch": 3.00398406374502,
+      "grad_norm": 0.2466438263654709,
+      "learning_rate": 2.4545313478339178e-06,
+      "loss": 0.5029895305633545,
+      "step": 754
+    },
+    {
+      "epoch": 3.0119521912350598,
+      "grad_norm": 0.3022938668727875,
+      "learning_rate": 2.4327557360651895e-06,
+      "loss": 0.5221859812736511,
+      "step": 756
+    },
+    {
+      "epoch": 3.0199203187250996,
+      "grad_norm": 1.0510525703430176,
+      "learning_rate": 2.411113452100211e-06,
+      "loss": 0.2634638845920563,
+      "step": 758
+    },
+    {
+      "epoch": 3.0278884462151394,
+      "grad_norm": 0.46127375960350037,
+      "learning_rate": 2.3896054366916365e-06,
+      "loss": 0.5368037819862366,
+      "step": 760
+    },
+    {
+      "epoch": 3.0358565737051793,
+      "grad_norm": 1.100422978401184,
+      "learning_rate": 2.3682326247557e-06,
+      "loss": 0.4069654047489166,
+      "step": 762
+    },
+    {
+      "epoch": 3.043824701195219,
+      "grad_norm": 0.5371790528297424,
+      "learning_rate": 2.346995945331576e-06,
+      "loss": 0.8391028642654419,
+      "step": 764
+    },
+    {
+      "epoch": 3.051792828685259,
+      "grad_norm": 0.43191012740135193,
+      "learning_rate": 2.325896321540993e-06,
+      "loss": 0.20137760043144226,
+      "step": 766
+    },
+    {
+      "epoch": 3.0597609561752988,
+      "grad_norm": 0.26182064414024353,
+      "learning_rate": 2.3049346705481086e-06,
+      "loss": 0.5045584440231323,
+      "step": 768
+    },
+    {
+      "epoch": 3.0677290836653386,
+      "grad_norm": 1.1320091485977173,
+      "learning_rate": 2.284111903519641e-06,
+      "loss": 0.6451742053031921,
+      "step": 770
+    },
+    {
+      "epoch": 3.0756972111553784,
+      "grad_norm": 0.2600367069244385,
+      "learning_rate": 2.26342892558526e-06,
+      "loss": 0.8277543783187866,
+      "step": 772
+    },
+    {
+      "epoch": 3.0836653386454183,
+      "grad_norm": 0.8583292961120605,
+      "learning_rate": 2.2428866357982515e-06,
+      "loss": 0.7643591165542603,
+      "step": 774
+    },
+    {
+      "epoch": 3.091633466135458,
+      "grad_norm": 0.2142353653907776,
+      "learning_rate": 2.222485927096424e-06,
+      "loss": 0.4709620773792267,
+      "step": 776
+    },
+    {
+      "epoch": 3.099601593625498,
+      "grad_norm": 0.5668942928314209,
+      "learning_rate": 2.2022276862633075e-06,
+      "loss": 0.494233638048172,
+      "step": 778
+    },
+    {
+      "epoch": 3.1075697211155378,
+      "grad_norm": 0.2683800756931305,
+      "learning_rate": 2.1821127938895938e-06,
+      "loss": 0.8124314546585083,
+      "step": 780
+    },
+    {
+      "epoch": 3.1155378486055776,
+      "grad_norm": 0.11633181571960449,
+      "learning_rate": 2.162142124334871e-06,
+      "loss": 0.2311871498823166,
+      "step": 782
+    },
+    {
+      "epoch": 3.1235059760956174,
+      "grad_norm": 0.26230666041374207,
+      "learning_rate": 2.1423165456896068e-06,
+      "loss": 0.8960323929786682,
+      "step": 784
+    },
+    {
+      "epoch": 3.1314741035856573,
+      "grad_norm": 0.21290867030620575,
+      "learning_rate": 2.1226369197374176e-06,
+      "loss": 0.6408299803733826,
+      "step": 786
+    },
+    {
+      "epoch": 3.139442231075697,
+      "grad_norm": 0.9733524918556213,
+      "learning_rate": 2.103104101917615e-06,
+      "loss": 0.45713627338409424,
+      "step": 788
+    },
+    {
+      "epoch": 3.147410358565737,
+      "grad_norm": 0.6192409992218018,
+      "learning_rate": 2.0837189412880044e-06,
+      "loss": 0.42875200510025024,
+      "step": 790
+    },
+    {
+      "epoch": 3.1553784860557768,
+      "grad_norm": 0.6196434497833252,
+      "learning_rate": 2.0644822804879964e-06,
+      "loss": 0.5648175477981567,
+      "step": 792
+    },
+    {
+      "epoch": 3.1633466135458166,
+      "grad_norm": 0.10330327600240707,
+      "learning_rate": 2.045394955701969e-06,
+      "loss": 0.15490427613258362,
+      "step": 794
+    },
+    {
+      "epoch": 3.1713147410358564,
+      "grad_norm": 2.709439992904663,
+      "learning_rate": 2.0264577966229222e-06,
+      "loss": 0.5055892467498779,
+      "step": 796
+    },
+    {
+      "epoch": 3.1792828685258963,
+      "grad_norm": 0.3505411446094513,
+      "learning_rate": 2.007671626416412e-06,
+      "loss": 0.517072319984436,
+      "step": 798
+    },
+    {
+      "epoch": 3.187250996015936,
+      "grad_norm": 0.485158771276474,
+      "learning_rate": 1.9890372616847663e-06,
+      "loss": 0.7315372228622437,
+      "step": 800
+    },
+    {
+      "epoch": 3.195219123505976,
+      "grad_norm": 0.19335967302322388,
+      "learning_rate": 1.9705555124315965e-06,
+      "loss": 0.5940592885017395,
+      "step": 802
+    },
+    {
+      "epoch": 3.2031872509960158,
+      "grad_norm": 1.2476129531860352,
+      "learning_rate": 1.95222718202658e-06,
+      "loss": 0.5281991362571716,
+      "step": 804
+    },
+    {
+      "epoch": 3.2111553784860556,
+      "grad_norm": 1.0794154405593872,
+      "learning_rate": 1.934053067170545e-06,
+      "loss": 0.7959721088409424,
+      "step": 806
+    },
+    {
+      "epoch": 3.2191235059760954,
+      "grad_norm": 0.6265828609466553,
+      "learning_rate": 1.9160339578608305e-06,
+      "loss": 0.16314499080181122,
+      "step": 808
+    },
+    {
+      "epoch": 3.2270916334661353,
+      "grad_norm": 0.20980319380760193,
+      "learning_rate": 1.8981706373569602e-06,
+      "loss": 0.5743886232376099,
+      "step": 810
+    },
+    {
+      "epoch": 3.235059760956175,
+      "grad_norm": 1.9075686931610107,
+      "learning_rate": 1.8804638821465821e-06,
+      "loss": 0.34441038966178894,
+      "step": 812
+    },
+    {
+      "epoch": 3.243027888446215,
+      "grad_norm": 0.21553963422775269,
+      "learning_rate": 1.8629144619117234e-06,
+      "loss": 0.6974723935127258,
+      "step": 814
+    },
+    {
+      "epoch": 3.2509960159362548,
+      "grad_norm": 1.038009762763977,
+      "learning_rate": 1.8455231394953348e-06,
+      "loss": 0.19147564470767975,
+      "step": 816
+    },
+    {
+      "epoch": 3.2589641434262946,
+      "grad_norm": 0.30175572633743286,
+      "learning_rate": 1.8282906708681238e-06,
+      "loss": 0.811612069606781,
+      "step": 818
+    },
+    {
+      "epoch": 3.2669322709163344,
+      "grad_norm": 0.6321881413459778,
+      "learning_rate": 1.811217805095699e-06,
+      "loss": 0.3769272267818451,
+      "step": 820
+    },
+    {
+      "epoch": 3.2749003984063743,
+      "grad_norm": 1.9237797260284424,
+      "learning_rate": 1.7943052843060133e-06,
+      "loss": 0.5534957051277161,
+      "step": 822
+    },
+    {
+      "epoch": 3.2828685258964145,
+      "grad_norm": 1.2198899984359741,
+      "learning_rate": 1.7775538436570937e-06,
+      "loss": 0.5237638354301453,
+      "step": 824
+    },
+    {
+      "epoch": 3.2908366533864544,
+      "grad_norm": 0.4041191041469574,
+      "learning_rate": 1.7609642113050996e-06,
+      "loss": 0.8719371557235718,
+      "step": 826
+    },
+    {
+      "epoch": 3.298804780876494,
+      "grad_norm": 0.2933843433856964,
+      "learning_rate": 1.7445371083726567e-06,
+      "loss": 0.8320538997650146,
+      "step": 828
+    },
+    {
+      "epoch": 3.306772908366534,
+      "grad_norm": 0.8950710296630859,
+      "learning_rate": 1.7282732489175224e-06,
+      "loss": 0.729328989982605,
+      "step": 830
+    },
+    {
+      "epoch": 3.314741035856574,
+      "grad_norm": 0.5004392862319946,
+      "learning_rate": 1.7121733399015406e-06,
+      "loss": 0.37355318665504456,
+      "step": 832
+    },
+    {
+      "epoch": 3.3227091633466137,
+      "grad_norm": 0.7134656310081482,
+      "learning_rate": 1.6962380811599138e-06,
+      "loss": 0.2689090669155121,
+      "step": 834
+    },
+    {
+      "epoch": 3.3306772908366535,
+      "grad_norm": 1.0973738431930542,
+      "learning_rate": 1.6804681653707816e-06,
+      "loss": 0.4466567039489746,
+      "step": 836
+    },
+    {
+      "epoch": 3.3386454183266934,
+      "grad_norm": 0.40353789925575256,
+      "learning_rate": 1.6648642780251096e-06,
+      "loss": 0.6369277238845825,
+      "step": 838
+    },
+    {
+      "epoch": 3.346613545816733,
+      "grad_norm": 0.9025564193725586,
+      "learning_rate": 1.6494270973968974e-06,
+      "loss": 0.1592463254928589,
+      "step": 840
+    },
+    {
+      "epoch": 3.354581673306773,
+      "grad_norm": 0.4777381122112274,
+      "learning_rate": 1.6341572945136902e-06,
+      "loss": 0.5088130235671997,
+      "step": 842
+    },
+    {
+      "epoch": 3.362549800796813,
+      "grad_norm": 0.2904771566390991,
+      "learning_rate": 1.6190555331274122e-06,
+      "loss": 0.5478823781013489,
+      "step": 844
+    },
+    {
+      "epoch": 3.3705179282868527,
+      "grad_norm": 0.28991320729255676,
+      "learning_rate": 1.6041224696855145e-06,
+      "loss": 0.8679755926132202,
+      "step": 846
+    },
+    {
+      "epoch": 3.3784860557768925,
+      "grad_norm": 0.23834329843521118,
+      "learning_rate": 1.589358753302439e-06,
+      "loss": 0.7919332385063171,
+      "step": 848
+    },
+    {
+      "epoch": 3.3864541832669324,
+      "grad_norm": 1.1245206594467163,
+      "learning_rate": 1.5747650257314068e-06,
+      "loss": 0.7111618518829346,
+      "step": 850
+    },
+    {
+      "epoch": 3.394422310756972,
+      "grad_norm": 0.2856602072715759,
+      "learning_rate": 1.5603419213365146e-06,
+      "loss": 0.8049482703208923,
+      "step": 852
+    },
+    {
+      "epoch": 3.402390438247012,
+      "grad_norm": 0.2575041949748993,
+      "learning_rate": 1.5460900670651705e-06,
+      "loss": 0.4878635108470917,
+      "step": 854
+    },
+    {
+      "epoch": 3.410358565737052,
+      "grad_norm": 0.4806252717971802,
+      "learning_rate": 1.5320100824208306e-06,
+      "loss": 0.7562990188598633,
+      "step": 856
+    },
+    {
+      "epoch": 3.4183266932270917,
+      "grad_norm": 0.28782302141189575,
+      "learning_rate": 1.51810257943608e-06,
+      "loss": 0.8691070079803467,
+      "step": 858
+    },
+    {
+      "epoch": 3.4262948207171315,
+      "grad_norm": 0.9343265295028687,
+      "learning_rate": 1.5043681626460196e-06,
+      "loss": 0.8567813038825989,
+      "step": 860
+    },
+    {
+      "epoch": 3.4342629482071714,
+      "grad_norm": 0.3011649549007416,
+      "learning_rate": 1.4908074290619972e-06,
+      "loss": 0.5705196261405945,
+      "step": 862
+    },
+    {
+      "epoch": 3.442231075697211,
+      "grad_norm": 0.6403683423995972,
+      "learning_rate": 1.477420968145652e-06,
+      "loss": 0.4003850221633911,
+      "step": 864
+    },
+    {
+      "epoch": 3.450199203187251,
+      "grad_norm": 0.22484008967876434,
+      "learning_rate": 1.4642093617832853e-06,
+      "loss": 0.6944388747215271,
+      "step": 866
+    },
+    {
+      "epoch": 3.458167330677291,
+      "grad_norm": 0.2931080162525177,
+      "learning_rate": 1.4511731842605805e-06,
+      "loss": 0.5801947116851807,
+      "step": 868
+    },
+    {
+      "epoch": 3.4661354581673307,
+      "grad_norm": 0.9420875310897827,
+      "learning_rate": 1.4383130022376298e-06,
+      "loss": 0.474017471075058,
+      "step": 870
+    },
+    {
+      "epoch": 3.4741035856573705,
+      "grad_norm": 0.0848192423582077,
+      "learning_rate": 1.4256293747243065e-06,
+      "loss": 0.4838365614414215,
+      "step": 872
+    },
+    {
+      "epoch": 3.4820717131474104,
+      "grad_norm": 0.1930120438337326,
+      "learning_rate": 1.4131228530559637e-06,
+      "loss": 0.5303343534469604,
+      "step": 874
+    },
+    {
+      "epoch": 3.49003984063745,
+      "grad_norm": 0.20566779375076294,
+      "learning_rate": 1.400793980869468e-06,
+      "loss": 0.4315550923347473,
+      "step": 876
+    },
+    {
+      "epoch": 3.49800796812749,
+      "grad_norm": 0.3322397470474243,
+      "learning_rate": 1.3886432940795744e-06,
+      "loss": 0.4921651780605316,
+      "step": 878
+    },
+    {
+      "epoch": 3.50597609561753,
+      "grad_norm": 0.12382643669843674,
+      "learning_rate": 1.3766713208556232e-06,
+      "loss": 0.3021814525127411,
+      "step": 880
+    },
+    {
+      "epoch": 3.5139442231075697,
+      "grad_norm": 0.4067063629627228,
+      "learning_rate": 1.3648785815985876e-06,
+      "loss": 0.42943093180656433,
+      "step": 882
+    },
+    {
+      "epoch": 3.5219123505976095,
+      "grad_norm": 1.1196112632751465,
+      "learning_rate": 1.353265588918448e-06,
+      "loss": 0.1019364520907402,
+      "step": 884
+    },
+    {
+      "epoch": 3.5298804780876494,
+      "grad_norm": 0.6129400730133057,
+      "learning_rate": 1.3418328476119136e-06,
+      "loss": 0.676003098487854,
+      "step": 886
+    },
+    {
+      "epoch": 3.537848605577689,
+      "grad_norm": 0.29097387194633484,
+      "learning_rate": 1.330580854640477e-06,
+      "loss": 0.7637168765068054,
+      "step": 888
+    },
+    {
+      "epoch": 3.545816733067729,
+      "grad_norm": 0.25648364424705505,
+      "learning_rate": 1.3195100991088142e-06,
+      "loss": 0.43930330872535706,
+      "step": 890
+    },
+    {
+      "epoch": 3.553784860557769,
+      "grad_norm": 0.9173100590705872,
+      "learning_rate": 1.308621062243524e-06,
+      "loss": 0.7101637721061707,
+      "step": 892
+    },
+    {
+      "epoch": 3.5617529880478087,
+      "grad_norm": 0.2515612840652466,
+      "learning_rate": 1.2979142173722054e-06,
+      "loss": 0.9359923005104065,
+      "step": 894
+    },
+    {
+      "epoch": 3.5697211155378485,
+      "grad_norm": 0.17758847773075104,
+      "learning_rate": 1.2873900299028896e-06,
+      "loss": 0.17805543541908264,
+      "step": 896
+    },
+    {
+      "epoch": 3.5776892430278884,
+      "grad_norm": 0.2228669822216034,
+      "learning_rate": 1.2770489573038073e-06,
+      "loss": 0.4978122413158417,
+      "step": 898
+    },
+    {
+      "epoch": 3.585657370517928,
+      "grad_norm": 0.3420407474040985,
+      "learning_rate": 1.2668914490834983e-06,
+      "loss": 0.1806340217590332,
+      "step": 900
+    },
+    {
+      "epoch": 3.593625498007968,
+      "grad_norm": 0.3964800238609314,
+      "learning_rate": 1.2569179467712805e-06,
+      "loss": 0.11508949846029282,
+      "step": 902
+    },
+    {
+      "epoch": 3.601593625498008,
+      "grad_norm": 0.27190253138542175,
+      "learning_rate": 1.2471288838980505e-06,
+      "loss": 0.5948676466941833,
+      "step": 904
+    },
+    {
+      "epoch": 3.6095617529880477,
+      "grad_norm": 0.5817465782165527,
+      "learning_rate": 1.237524685977444e-06,
+      "loss": 0.4821530878543854,
+      "step": 906
+    },
+    {
+      "epoch": 3.6175298804780875,
+      "grad_norm": 0.3387485444545746,
+      "learning_rate": 1.2281057704873331e-06,
+      "loss": 0.7491830587387085,
+      "step": 908
+    },
+    {
+      "epoch": 3.6254980079681274,
+      "grad_norm": 0.4753982126712799,
+      "learning_rate": 1.2188725468516876e-06,
+      "loss": 0.7225444316864014,
+      "step": 910
+    },
+    {
+      "epoch": 3.633466135458167,
+      "grad_norm": 0.12187855690717697,
+      "learning_rate": 1.2098254164227706e-06,
+      "loss": 0.03353629261255264,
+      "step": 912
+    },
+    {
+      "epoch": 3.641434262948207,
+      "grad_norm": 0.34467142820358276,
+      "learning_rate": 1.200964772463695e-06,
+      "loss": 0.6757269501686096,
+      "step": 914
+    },
+    {
+      "epoch": 3.649402390438247,
+      "grad_norm": 0.603409469127655,
+      "learning_rate": 1.1922910001313312e-06,
+      "loss": 0.6553897261619568,
+      "step": 916
+    },
+    {
+      "epoch": 3.6573705179282867,
+      "grad_norm": 0.5768171548843384,
+      "learning_rate": 1.1838044764595617e-06,
+      "loss": 0.12231604754924774,
+      "step": 918
+    },
+    {
+      "epoch": 3.6653386454183265,
+      "grad_norm": 0.26420044898986816,
+      "learning_rate": 1.1755055703428958e-06,
+      "loss": 0.8007838726043701,
+      "step": 920
+    },
+    {
+      "epoch": 3.6733067729083664,
+      "grad_norm": 0.5795018672943115,
+      "learning_rate": 1.1673946425204288e-06,
+      "loss": 0.10008751600980759,
+      "step": 922
+    },
+    {
+      "epoch": 3.681274900398406,
+      "grad_norm": 0.5196491479873657,
+      "learning_rate": 1.1594720455601664e-06,
+      "loss": 0.8696997761726379,
+      "step": 924
+    },
+    {
+      "epoch": 3.6892430278884465,
+      "grad_norm": 0.2617645561695099,
+      "learning_rate": 1.1517381238436986e-06,
+      "loss": 0.23366229236125946,
+      "step": 926
+    },
+    {
+      "epoch": 3.6972111553784863,
+      "grad_norm": 0.28648459911346436,
+      "learning_rate": 1.1441932135512262e-06,
+      "loss": 0.45714905858039856,
+      "step": 928
+    },
+    {
+      "epoch": 3.705179282868526,
+      "grad_norm": 0.5338141322135925,
+      "learning_rate": 1.1368376426469526e-06,
+      "loss": 0.6726741194725037,
+      "step": 930
+    },
+    {
+      "epoch": 3.713147410358566,
+      "grad_norm": 0.46902549266815186,
+      "learning_rate": 1.1296717308648237e-06,
+      "loss": 0.8055936694145203,
+      "step": 932
+    },
+    {
+      "epoch": 3.721115537848606,
+      "grad_norm": 0.9705061316490173,
+      "learning_rate": 1.1226957896946332e-06,
+      "loss": 0.4154506325721741,
+      "step": 934
+    },
+    {
+      "epoch": 3.7290836653386457,
+      "grad_norm": 0.2266857773065567,
+      "learning_rate": 1.115910122368479e-06,
+      "loss": 0.7319180965423584,
+      "step": 936
+    },
+    {
+      "epoch": 3.7370517928286855,
+      "grad_norm": 0.9299195408821106,
+      "learning_rate": 1.1093150238475847e-06,
+      "loss": 0.38912248611450195,
+      "step": 938
+    },
+    {
+      "epoch": 3.7450199203187253,
+      "grad_norm": 0.4368281662464142,
+      "learning_rate": 1.1029107808094789e-06,
+      "loss": 0.6652960777282715,
+      "step": 940
+    },
+    {
+      "epoch": 3.752988047808765,
+      "grad_norm": 1.029328465461731,
+      "learning_rate": 1.0966976716355294e-06,
+      "loss": 0.539993941783905,
+      "step": 942
+    },
+    {
+      "epoch": 3.760956175298805,
+      "grad_norm": 0.23156608641147614,
+      "learning_rate": 1.0906759663988475e-06,
+      "loss": 0.4666006565093994,
+      "step": 944
+    },
+    {
+      "epoch": 3.768924302788845,
+      "grad_norm": 0.2646404504776001,
+      "learning_rate": 1.084845926852547e-06,
+      "loss": 0.8639662265777588,
+      "step": 946
+    },
+    {
+      "epoch": 3.7768924302788847,
+      "grad_norm": 0.2451619654893875,
+      "learning_rate": 1.0792078064183652e-06,
+      "loss": 0.45361295342445374,
+      "step": 948
+    },
+    {
+      "epoch": 3.7848605577689245,
+      "grad_norm": 0.3057004511356354,
+      "learning_rate": 1.0737618501756455e-06,
+      "loss": 0.8175570368766785,
+      "step": 950
+    },
+    {
+      "epoch": 3.7928286852589643,
+      "grad_norm": 0.3106706142425537,
+      "learning_rate": 1.0685082948506875e-06,
+      "loss": 0.8223646879196167,
+      "step": 952
+    },
+    {
+      "epoch": 3.800796812749004,
+      "grad_norm": 0.2798798382282257,
+      "learning_rate": 1.063447368806457e-06,
+      "loss": 0.43498510122299194,
+      "step": 954
+    },
+    {
+      "epoch": 3.808764940239044,
+      "grad_norm": 0.5249965786933899,
+      "learning_rate": 1.0585792920326565e-06,
+      "loss": 0.627868115901947,
+      "step": 956
+    },
+    {
+      "epoch": 3.816733067729084,
+      "grad_norm": 0.6519578099250793,
+      "learning_rate": 1.0539042761361646e-06,
+      "loss": 0.42182379961013794,
+      "step": 958
+    },
+    {
+      "epoch": 3.8247011952191237,
+      "grad_norm": 0.28383609652519226,
+      "learning_rate": 1.0494225243318373e-06,
+      "loss": 0.5007190108299255,
+      "step": 960
+    },
+    {
+      "epoch": 3.8326693227091635,
+      "grad_norm": 0.1351151019334793,
+      "learning_rate": 1.0451342314336755e-06,
+      "loss": 0.4886111915111542,
+      "step": 962
+    },
+    {
+      "epoch": 3.8406374501992033,
+      "grad_norm": 0.39173927903175354,
+      "learning_rate": 1.0410395838463547e-06,
+      "loss": 0.8553838729858398,
+      "step": 964
+    },
+    {
+      "epoch": 3.848605577689243,
+      "grad_norm": 0.43374907970428467,
+      "learning_rate": 1.0371387595571248e-06,
+      "loss": 0.286433607339859,
+      "step": 966
+    },
+    {
+      "epoch": 3.856573705179283,
+      "grad_norm": 0.2957891821861267,
+      "learning_rate": 1.0334319281280715e-06,
+      "loss": 0.5115336179733276,
+      "step": 968
+    },
+    {
+      "epoch": 3.864541832669323,
+      "grad_norm": 0.5552667379379272,
+      "learning_rate": 1.0299192506887467e-06,
+      "loss": 0.5729485154151917,
+      "step": 970
+    },
+    {
+      "epoch": 3.8725099601593627,
+      "grad_norm": 0.0714365616440773,
+      "learning_rate": 1.0266008799291635e-06,
+      "loss": 0.4521854519844055,
+      "step": 972
+    },
+    {
+      "epoch": 3.8804780876494025,
+      "grad_norm": 0.10103405267000198,
+      "learning_rate": 1.0234769600931612e-06,
+      "loss": 0.37017208337783813,
+      "step": 974
+    },
+    {
+      "epoch": 3.8884462151394423,
+      "grad_norm": 0.2354981154203415,
+      "learning_rate": 1.0205476269721312e-06,
+      "loss": 0.8241890072822571,
+      "step": 976
+    },
+    {
+      "epoch": 3.896414342629482,
+      "grad_norm": 1.130471110343933,
+      "learning_rate": 1.0178130078991196e-06,
+      "loss": 0.4780001640319824,
+      "step": 978
+    },
+    {
+      "epoch": 3.904382470119522,
+      "grad_norm": 0.3868483603000641,
+      "learning_rate": 1.0152732217432875e-06,
+      "loss": 0.3425574004650116,
+      "step": 980
+    },
+    {
+      "epoch": 3.912350597609562,
+      "grad_norm": 0.4195859432220459,
+      "learning_rate": 1.0129283789047473e-06,
+      "loss": 0.3609636425971985,
+      "step": 982
+    },
+    {
+      "epoch": 3.9203187250996017,
+      "grad_norm": 0.28839045763015747,
+      "learning_rate": 1.0107785813097621e-06,
+      "loss": 0.7903327941894531,
+      "step": 984
+    },
+    {
+      "epoch": 3.9282868525896415,
+      "grad_norm": 0.4476912319660187,
+      "learning_rate": 1.0088239224063153e-06,
+      "loss": 0.17196184396743774,
+      "step": 986
+    },
+    {
+      "epoch": 3.9362549800796813,
+      "grad_norm": 0.6139546632766724,
+      "learning_rate": 1.0070644871600495e-06,
+      "loss": 0.5698275566101074,
+      "step": 988
+    },
+    {
+      "epoch": 3.944223107569721,
+      "grad_norm": 0.07818426936864853,
+      "learning_rate": 1.0055003520505714e-06,
+      "loss": 0.45491334795951843,
+      "step": 990
+    },
+    {
+      "epoch": 3.952191235059761,
+      "grad_norm": 0.3576550781726837,
+      "learning_rate": 1.00413158506813e-06,
+      "loss": 0.8213901519775391,
+      "step": 992
+    },
+    {
+      "epoch": 3.960159362549801,
+      "grad_norm": 0.08875592797994614,
+      "learning_rate": 1.0029582457106588e-06,
+      "loss": 0.19506685435771942,
+      "step": 994
+    },
+    {
+      "epoch": 3.9681274900398407,
+      "grad_norm": 0.3288118541240692,
+      "learning_rate": 1.00198038498119e-06,
+      "loss": 0.7251580953598022,
+      "step": 996
+    },
+    {
+      "epoch": 3.9760956175298805,
+      "grad_norm": 0.25357329845428467,
+      "learning_rate": 1.0011980453856391e-06,
+      "loss": 0.4798094630241394,
+      "step": 998
+    },
+    {
+      "epoch": 3.9840637450199203,
+      "grad_norm": 0.28475239872932434,
+      "learning_rate": 1.0006112609309548e-06,
+      "loss": 0.359422504901886,
+      "step": 1000
+    },
+    {
+      "epoch": 3.99203187250996,
+      "grad_norm": 0.3201379179954529,
+      "learning_rate": 1.0002200571236429e-06,
+      "loss": 0.8232021927833557,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.2544263005256653,
+      "learning_rate": 1.0000244509686564e-06,
+      "loss": 0.21048825979232788,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0,
+      "step": 1004,
+      "total_flos": 4.038502240003031e+18,
+      "train_loss": 0.8272027168051893,
+      "train_runtime": 10128.8027,
+      "train_samples_per_second": 5.947,
+      "train_steps_per_second": 0.099
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1004,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 99999,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.038502240003031e+18,
+  "train_batch_size": 3,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

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