Upload folder using huggingface_hub

LICENSE.md

@@ -0,0 +1,36 @@
+STABLECODE RESEARCH LICENSE AGREEMENT	
+Dated: August 8, 2023
+
+"Agreement" means the terms and conditions for use, reproduction, distribution and modification of the Software Products set forth herein.
+
+"Documentation" means any specifications, manuals, documentation, and other written information provided by Stability AI related to the Software.
+
+"Licensee" or "you" means you, or your employer or any other person or entity (if you are entering into this Agreement on such person or entity's behalf), of the age required under applicable laws, rules or regulations to provide legal consent and that has legal authority to bind your employer or such other person or entity if you are entering in this Agreement on their behalf.
+
+"Stability AI" or "we" means Stability AI Ltd. 
+
+"Software" means, collectively, Stability AI’s proprietary StableCode made available under this Agreement.
+
+"Software Products" means Software and Documentation. 
+
+By using or distributing any portion or element of the Software Products, you agree to be bound by this Agreement.
+
+1. License Rights and Redistribution.
+
+a.  Subject to your compliance with this Agreement and the Documentation, Stability AI grants you a non-exclusive, worldwide, non-transferable, non-sublicensable, revocable, royalty free and limited license under Stability AI’s intellectual property or other rights owned by Stability AI embodied in the Software Products to reproduce, distribute, and create derivative works of the Software Products solely for your non-commercial research purposes.
+
+b.	If you distribute or make the Software Products, or any derivative works thereof, available to a third party, you shall (i) provide a copy of this Agreement to such third party, and (ii) retain the following attribution notice within a "Notice" text file distributed as a part of such copies: "StableCode is licensed under the StableCode Research License, Copyright (c) Stability AI Ltd. All Rights Reserved.”
+
+2. Disclaimer of Warranty. UNLESS REQUIRED BY APPLICABLE LAW, THE SOFTWARE PRODUCTS  AND ANY OUTPUT AND RESULTS THEREFROM ARE PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, ANY WARRANTIES OF TITLE, NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE. YOU ARE SOLELY RESPONSIBLE FOR DETERMINING THE APPROPRIATENESS OF USING OR REDISTRIBUTING THE SOFTWARE PRODUCTS AND ASSUME ANY RISKS ASSOCIATED WITH YOUR USE OF THE SOFTWARE PRODUCTS AND ANY OUTPUT AND RESULTS. 
+
+3. Limitation of Liability. IN NO EVENT WILL STABILITY AI OR ITS AFFILIATES BE LIABLE UNDER ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, TORT, NEGLIGENCE, PRODUCTS LIABILITY, OR OTHERWISE, ARISING OUT OF THIS AGREEMENT, FOR ANY LOST PROFITS OR ANY INDIRECT, SPECIAL, CONSEQUENTIAL, INCIDENTAL, EXEMPLARY OR PUNITIVE DAMAGES, EVEN IF STABILITY AI OR ITS AFFILIATES HAVE BEEN ADVISED OF THE POSSIBILITY OF ANY OF THE FOREGOING. 
+
+4. Intellectual Property.
+
+a. 	 No trademark licenses are granted under this Agreement, and in connection with the Software Products, neither Stability AI nor Licensee may use any name or mark owned by or associated with the other or any of its affiliates, except as required for reasonable and customary use in describing and redistributing the Software Products. 
+
+b.   Subject to Stability AI’s ownership of the Software Products and derivatives made by or for Stability AI, with respect to any derivative works and modifications of the Software Products that are made by you, as between you and Stability AI, you are and will be the owner of such derivative works and modifications. 
+
+c.   If you institute litigation or other proceedings against Stability AI (including a cross-claim or counterclaim in a lawsuit) alleging that the Software Products or associated outputs or results, or any portion of any of the foregoing, constitutes infringement of intellectual property or other rights owned or licensable by you, then any licenses granted to you under this Agreement shall terminate as of the date such litigation or claim is filed or instituted. You will indemnify and hold harmless Stability AI from and against any claim by any third party arising out of or related to your use or distribution of the Software Products in violation of this Agreement. 
+
+5.   Term and Termination. The term of this Agreement will commence upon your acceptance of this Agreement or access to the Software Products and will continue in full force and effect until terminated in accordance with the terms and conditions herein. Stability AI may terminate this Agreement if you are in breach of any term or condition of this Agreement. Upon termination of this Agreement, you shall delete and cease use of the Software Products. Sections 2-4 shall survive the termination of this Agreement. 

NOTICE.md

@@ -0,0 +1 @@
+StableCode is licensed under the StableCode Research License, Copyright (c) Stability AI Ltd. All Rights Reserved

README.md

@@ -0,0 +1,139 @@
+---
+language:
+- code
+tags:
+- causal-lm
+model-index:
+  - name: stabilityai/stablecode-instruct-alpha-3b
+    results:
+      - task:
+          type: text-generation
+        dataset:
+          type: openai_humaneval
+          name: HumanEval
+        metrics:
+          - name: pass@1
+            type: pass@1
+            value: 0.2689	
+            verified: false
+          - name: pass@10
+            type: pass@10
+            value: 0.3618
+            verified: false
+
+license: other
+extra_gated_prompt: >
+  STABLECODE RESEARCH LICENSE AGREEMENT	
+  Dated: August 8, 2023
+  
+  "Agreement" means the terms and conditions for use, reproduction, distribution and modification of the Software Products set forth herein.
+  
+  "Documentation" means any specifications, manuals, documentation, and other written information provided by Stability AI related to the Software.
+  
+  "Licensee" or "you" means you, or your employer or any other person or entity (if you are entering into this Agreement on such person or entity's behalf), of the age required under applicable laws, rules or regulations to provide legal consent and that has legal authority to bind your employer or such other person or entity if you are entering in this Agreement on their behalf.
+  
+  "Stability AI" or "we" means Stability AI Ltd. 
+  
+  "Software" means, collectively, Stability AI’s proprietary StableCode made available under this Agreement.
+  
+  "Software Products" means Software and Documentation. 
+  
+  By using or distributing any portion or element of the Software Products, you agree to be bound by this Agreement.
+  
+  1. License Rights and Redistribution.
+  
+  a.  Subject to your compliance with this Agreement and the Documentation, Stability AI grants you a non-exclusive, worldwide, non-transferable, non-sublicensable, revocable, royalty free and limited license under Stability AI’s intellectual property or other rights owned by Stability AI embodied in the Software Products to reproduce, distribute, and create derivative works of the Software Products solely for your non-commercial research purposes.
+  
+  b.	If you distribute or make the Software Products, or any derivative works thereof, available to a third party, you shall (i) provide a copy of this Agreement to such third party, and (ii) retain the following attribution notice within a "Notice" text file distributed as a part of such copies: "StableCode is licensed under the StableCode Research License, Copyright (c) Stability AI Ltd. All Rights Reserved.”
+  
+  2. Disclaimer of Warranty. UNLESS REQUIRED BY APPLICABLE LAW, THE SOFTWARE PRODUCTS  AND ANY OUTPUT AND RESULTS THEREFROM ARE PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, ANY WARRANTIES OF TITLE, NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE. YOU ARE SOLELY RESPONSIBLE FOR DETERMINING THE APPROPRIATENESS OF USING OR REDISTRIBUTING THE SOFTWARE PRODUCTS AND ASSUME ANY RISKS ASSOCIATED WITH YOUR USE OF THE SOFTWARE PRODUCTS AND ANY OUTPUT AND RESULTS. 
+  
+  3. Limitation of Liability. IN NO EVENT WILL STABILITY AI OR ITS AFFILIATES BE LIABLE UNDER ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, TORT, NEGLIGENCE, PRODUCTS LIABILITY, OR OTHERWISE, ARISING OUT OF THIS AGREEMENT, FOR ANY LOST PROFITS OR ANY INDIRECT, SPECIAL, CONSEQUENTIAL, INCIDENTAL, EXEMPLARY OR PUNITIVE DAMAGES, EVEN IF STABILITY AI OR ITS AFFILIATES HAVE BEEN ADVISED OF THE POSSIBILITY OF ANY OF THE FOREGOING. 
+  
+  4. Intellectual Property.
+  
+  a. 	 No trademark licenses are granted under this Agreement, and in connection with the Software Products, neither Stability AI nor Licensee may use any name or mark owned by or associated with the other or any of its affiliates, except as required for reasonable and customary use in describing and redistributing the Software Products. 
+  
+  b.   Subject to Stability AI’s ownership of the Software Products and derivatives made by or for Stability AI, with respect to any derivative works and modifications of the Software Products that are made by you, as between you and Stability AI, you are and will be the owner of such derivative works and modifications. 
+  
+  c.   If you institute litigation or other proceedings against Stability AI (including a cross-claim or counterclaim in a lawsuit) alleging that the Software Products or associated outputs or results, or any portion of any of the foregoing, constitutes infringement of intellectual property or other rights owned or licensable by you, then any licenses granted to you under this Agreement shall terminate as of the date such litigation or claim is filed or instituted. You will indemnify and hold harmless Stability AI from and against any claim by any third party arising out of or related to your use or distribution of the Software Products in violation of this Agreement. 
+  
+  5.   Term and Termination. The term of this Agreement will commence upon your acceptance of this Agreement or access to the Software Products and will continue in full force and effect until terminated in accordance with the terms and conditions herein. Stability AI may terminate this Agreement if you are in breach of any term or condition of this Agreement. Upon termination of this Agreement, you shall delete and cease use of the Software Products. Sections 2-4 shall survive the termination of this Agreement.
+extra_gated_fields:
+ # Company: text
+ # Country: text
+ I agree to use this model for research use ONLY: checkbox
+
+---
+# `StableCode-Instruct-Alpha-3B`
+
+## Model Description
+
+`StableCode-Instruct-Alpha-3B` is a 3 billion parameter decoder-only instruction tuned code model pre-trained on diverse set of programming languages that topped the stackoverflow developer survey.
+
+## Usage
+The model is intended to follow instruction to generate code. The dataset used to train the model is formatted in Alpaca format.
+Get started generating code with `StableCode-Instruct-Alpha-3B` by using the following code snippet:
+
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+tokenizer = AutoTokenizer.from_pretrained("stabilityai/stablecode-instruct-alpha-3b")
+model = AutoModelForCausalLM.from_pretrained(
+  "stabilityai/stablecode-instruct-alpha-3b",
+  trust_remote_code=True,
+  torch_dtype="auto",
+)
+model.cuda()
+inputs = tokenizer("###Instruction\nGenerate a python function to find number of CPU cores###Response\n", return_tensors="pt").to("cuda")
+tokens = model.generate(
+  **inputs,
+  max_new_tokens=48,
+  temperature=0.2,
+  do_sample=True,
+)
+print(tokenizer.decode(tokens[0], skip_special_tokens=True))
+```
+
+## Model Details
+
+* **Developed by**: [Stability AI](https://stability.ai/)
+* **Model type**: `StableCode-Instruct-Alpha-3B` models are auto-regressive language models based on the transformer decoder architecture.
+* **Language(s)**: Code
+* **Library**: [GPT-NeoX](https://github.com/EleutherAI/gpt-neox)
+* **License** : Model checkpoints are licensed under the [StableCode Research License](https://huggingface.co/stabilityai/stablecode-instruct-alpha-3b/blob/main/LICENSE.md) Copyright (c) Stability AI Ltd. All Rights Reserved
+* **Contact**: For questions and comments about the model, please email `lm@stability.ai`
+
+### Model Architecture
+
+| Parameters     | Hidden Size | Layers | Heads | Sequence Length |
+|----------------|-------------|--------|-------|-----------------|
+| 2,796,431,360  | 2560        | 32     | 32    | 4096            |
+
+
+* **Decoder Layer**: Parallel Attention and MLP residuals with a single input LayerNorm ([Wang & Komatsuzaki, 2021](https://github.com/kingoflolz/mesh-transformer-jax/tree/master))
+* **Position Embeddings**: Rotary Position Embeddings ([Su et al., 2021](https://arxiv.org/abs/2104.09864))
+* **Bias**: LayerNorm bias terms only
+
+## Training
+
+`StableCode-Instruct-Alpha-3B` is the instruction finetuned version on [StableCode-Completion-Alpha-3B](https://huggingface.co/stabilityai/stablecode-completion-alpha-3b) with code instruction datasets.
+
+## Use and Limitations
+
+### Intended Use
+
+StableCode-Instruct-Alpha-3B independently generates new code completions, but we recommend that you use StableCode-Instruct-Alpha-3B together with the tool developed by BigCode and HuggingFace [(huggingface/huggingface-vscode: Code completion VSCode extension for OSS models (github.com))](https://github.com/huggingface/huggingface-vscode), to identify and, if necessary, attribute any outputs that match training code.
+
+### Limitations and bias
+
+This model is intended to be used responsibly. It is not intended to be used to create unlawful content of any kind, to further any unlawful activity, or to engage in activities with a high risk of physical or economic harm.
+
+## How to cite
+
+```bibtex
+@misc{StableCodeInstructAlpha, 
+      url={[https://huggingface.co/stabilityai/stablecode-instruct-alpha-3b](https://huggingface.co/stabilityai/stablecode-instruct-alpha-3b)}, 
+      title={Stable Code Instruct Alpha}, 
+      author={Adithyan, Reshinth and Phung, Duy and Cooper, Nathan and Pinnaparaju, Nikhil and Laforte, Christian}
+}
+```

config.json

@@ -0,0 +1,26 @@
+{
+  "_name_or_path": "/fsx/ckpts/3b_tok=starcoder_data=star_coder_model=stable-code-no-fim-specific-langs/global_step150000_hf",
+  "architectures": [
+    "GPTNeoXForCausalLM"
+  ],
+  "bos_token_id": 0,
+  "classifier_dropout": 0.1,
+  "eos_token_id": 0,
+  "hidden_act": "gelu",
+  "hidden_size": 2560,
+  "initializer_range": 0.02,
+  "intermediate_size": 10240,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 4096,
+  "model_type": "gpt_neox",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "rotary_emb_base": 10000,
+  "rotary_pct": 0.25,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.30.2",
+  "use_cache": false,
+  "use_parallel_residual": true,
+  "vocab_size": 49152
+}

ctranslate2/config.json

@@ -0,0 +1,6 @@
+{
+  "bos_token": null,
+  "eos_token": "<|endoftext|>",
+  "layer_norm_epsilon": null,
+  "unk_token": null
+}

ctranslate2/model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6e02c248f9f5aeeb6cf1da4e2d3672a6054d7a56bba7cdb5a9c0aef4abff534
+size 5538646818

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 0,
+  "eos_token_id": 0,
+  "transformers_version": "4.30.2"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7674c4fee28a3bca6cdb6950ad03719b6932845873e169ab2cb26420f91b74ce
+size 6075550304

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4ec4a192da28265510f0407bccbc0205c1b8293488523ca84ecb0255952c4fd8
+size 6075650877

special_tokens_map.json

@@ -0,0 +1,4 @@
+{
+  "eos_token": "<|endoftext|>",
+  "pad_token": "<|endoftext|>"
+}

tabby.json

@@ -0,0 +1,3 @@
+{
+    "auto_model": "AutoModelForCausalLM"
+}

tokenizer_config.json

@@ -0,0 +1,6 @@
+{
+  "clean_up_tokenization_spaces": true,
+  "model_max_length": 2048,
+  "padding_side": "right",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}

trainer_state.json

@@ -0,0 +1,4327 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "global_step": 717,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 0.7678,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.7731,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.8476,
+      "step": 3
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.783,
+      "step": 4
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.7332,
+      "step": 5
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.6563,
+      "step": 6
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.6574,
+      "step": 7
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.6471,
+      "step": 8
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.5971,
+      "step": 9
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.5837,
+      "step": 10
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.5884,
+      "step": 11
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 0.5916,
+      "step": 12
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.181818181818182e-05,
+      "loss": 0.5759,
+      "step": 13
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 0.5464,
+      "step": 14
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.5675,
+      "step": 15
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 0.5477,
+      "step": 16
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5454545454545454e-05,
+      "loss": 0.5162,
+      "step": 17
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 0.5018,
+      "step": 18
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7272727272727274e-05,
+      "loss": 0.5149,
+      "step": 19
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.5485,
+      "step": 20
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9090909090909094e-05,
+      "loss": 0.5174,
+      "step": 21
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 0.5476,
+      "step": 22
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999897835644166e-05,
+      "loss": 0.5322,
+      "step": 23
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999591344664163e-05,
+      "loss": 0.4921,
+      "step": 24
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999080533322486e-05,
+      "loss": 0.5514,
+      "step": 25
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998365412056476e-05,
+      "loss": 0.4966,
+      "step": 26
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999744599547812e-05,
+      "loss": 0.4834,
+      "step": 27
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999632230237373e-05,
+      "loss": 0.5289,
+      "step": 28
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999499435570359e-05,
+      "loss": 0.4914,
+      "step": 29
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999346218260146e-05,
+      "loss": 0.5273,
+      "step": 30
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999172581437403e-05,
+      "loss": 0.4999,
+      "step": 31
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9989785286500294e-05,
+      "loss": 0.4953,
+      "step": 32
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9987640638630812e-05,
+      "loss": 0.4888,
+      "step": 33
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998529191458689e-05,
+      "loss": 0.4968,
+      "step": 34
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9982739162359707e-05,
+      "loss": 0.4915,
+      "step": 35
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.997998243410932e-05,
+      "loss": 0.4958,
+      "step": 36
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.99770217861636e-05,
+      "loss": 0.5136,
+      "step": 37
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9973857279017092e-05,
+      "loss": 0.4989,
+      "step": 38
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9970488977329757e-05,
+      "loss": 0.4797,
+      "step": 39
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996691694992567e-05,
+      "loss": 0.4806,
+      "step": 40
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9963141269791606e-05,
+      "loss": 0.4641,
+      "step": 41
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9959162014075553e-05,
+      "loss": 0.4734,
+      "step": 42
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.995497926408513e-05,
+      "loss": 0.4716,
+      "step": 43
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9950593105285927e-05,
+      "loss": 0.4856,
+      "step": 44
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9946003627299766e-05,
+      "loss": 0.4658,
+      "step": 45
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9941210923902853e-05,
+      "loss": 0.4677,
+      "step": 46
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9936215093023884e-05,
+      "loss": 0.4633,
+      "step": 47
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9931016236742026e-05,
+      "loss": 0.4699,
+      "step": 48
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.992561446128484e-05,
+      "loss": 0.44,
+      "step": 49
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9920009877026106e-05,
+      "loss": 0.4759,
+      "step": 50
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9914202598483576e-05,
+      "loss": 0.4693,
+      "step": 51
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.990819274431662e-05,
+      "loss": 0.4647,
+      "step": 52
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9901980437323818e-05,
+      "loss": 0.4502,
+      "step": 53
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9895565804440435e-05,
+      "loss": 0.5081,
+      "step": 54
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9888948976735843e-05,
+      "loss": 0.4724,
+      "step": 55
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9882130089410822e-05,
+      "loss": 0.4941,
+      "step": 56
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9875109281794828e-05,
+      "loss": 0.4827,
+      "step": 57
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.986788669734311e-05,
+      "loss": 0.4633,
+      "step": 58
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.986046248363381e-05,
+      "loss": 0.4893,
+      "step": 59
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985283679236493e-05,
+      "loss": 0.4672,
+      "step": 60
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9845009779351235e-05,
+      "loss": 0.4666,
+      "step": 61
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9836981604521077e-05,
+      "loss": 0.4673,
+      "step": 62
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9828752431913116e-05,
+      "loss": 0.4754,
+      "step": 63
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9820322429672978e-05,
+      "loss": 0.5032,
+      "step": 64
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9811691770049806e-05,
+      "loss": 0.4799,
+      "step": 65
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9802860629392765e-05,
+      "loss": 0.4699,
+      "step": 66
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9793829188147406e-05,
+      "loss": 0.472,
+      "step": 67
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9784597630852008e-05,
+      "loss": 0.5005,
+      "step": 68
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.97751661461338e-05,
+      "loss": 0.4388,
+      "step": 69
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9765534926705082e-05,
+      "loss": 0.4659,
+      "step": 70
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.975570416935932e-05,
+      "loss": 0.4816,
+      "step": 71
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.974567407496712e-05,
+      "loss": 0.4659,
+      "step": 72
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9735444848472108e-05,
+      "loss": 0.472,
+      "step": 73
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9725016698886748e-05,
+      "loss": 0.468,
+      "step": 74
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9714389839288073e-05,
+      "loss": 0.4568,
+      "step": 75
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9703564486813342e-05,
+      "loss": 0.4627,
+      "step": 76
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9692540862655587e-05,
+      "loss": 0.497,
+      "step": 77
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.96813191920591e-05,
+      "loss": 0.496,
+      "step": 78
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9669899704314828e-05,
+      "loss": 0.4655,
+      "step": 79
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9658282632755694e-05,
+      "loss": 0.4611,
+      "step": 80
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.964646821475183e-05,
+      "loss": 0.4795,
+      "step": 81
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9634456691705705e-05,
+      "loss": 0.4558,
+      "step": 82
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9622248309047233e-05,
+      "loss": 0.476,
+      "step": 83
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.960984331622872e-05,
+      "loss": 0.4417,
+      "step": 84
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.959724196671978e-05,
+      "loss": 0.4682,
+      "step": 85
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9584444518002178e-05,
+      "loss": 0.485,
+      "step": 86
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9571451231564523e-05,
+      "loss": 0.4674,
+      "step": 87
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.955826237289697e-05,
+      "loss": 0.4464,
+      "step": 88
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9544878211485763e-05,
+      "loss": 0.4677,
+      "step": 89
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9531299020807752e-05,
+      "loss": 0.4343,
+      "step": 90
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9517525078324787e-05,
+      "loss": 0.4506,
+      "step": 91
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9503556665478066e-05,
+      "loss": 0.4669,
+      "step": 92
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9489394067682365e-05,
+      "loss": 0.4587,
+      "step": 93
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9475037574320217e-05,
+      "loss": 0.4337,
+      "step": 94
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.946048747873601e-05,
+      "loss": 0.4547,
+      "step": 95
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9445744078229967e-05,
+      "loss": 0.4605,
+      "step": 96
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9430807674052092e-05,
+      "loss": 0.4508,
+      "step": 97
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9415678571396006e-05,
+      "loss": 0.4642,
+      "step": 98
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9400357079392714e-05,
+      "loss": 0.4528,
+      "step": 99
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9384843511104294e-05,
+      "loss": 0.4679,
+      "step": 100
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.936913818351748e-05,
+      "loss": 0.4425,
+      "step": 101
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9353241417537216e-05,
+      "loss": 0.4758,
+      "step": 102
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.933715353798006e-05,
+      "loss": 0.4516,
+      "step": 103
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9320874873567598e-05,
+      "loss": 0.4519,
+      "step": 104
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.930440575691967e-05,
+      "loss": 0.4861,
+      "step": 105
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9287746524547627e-05,
+      "loss": 0.4635,
+      "step": 106
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9270897516847406e-05,
+      "loss": 0.4607,
+      "step": 107
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9253859078092616e-05,
+      "loss": 0.4538,
+      "step": 108
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.923663155642748e-05,
+      "loss": 0.4647,
+      "step": 109
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9219215303859732e-05,
+      "loss": 0.4524,
+      "step": 110
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9201610676253412e-05,
+      "loss": 0.4296,
+      "step": 111
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9183818033321612e-05,
+      "loss": 0.4378,
+      "step": 112
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.916583773861911e-05,
+      "loss": 0.4793,
+      "step": 113
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9147670159534953e-05,
+      "loss": 0.4423,
+      "step": 114
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.912931566728494e-05,
+      "loss": 0.4824,
+      "step": 115
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9110774636904052e-05,
+      "loss": 0.4399,
+      "step": 116
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9092047447238775e-05,
+      "loss": 0.478,
+      "step": 117
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9073134480939353e-05,
+      "loss": 0.4661,
+      "step": 118
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9054036124452e-05,
+      "loss": 0.4806,
+      "step": 119
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9034752768010965e-05,
+      "loss": 0.4437,
+      "step": 120
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.901528480563059e-05,
+      "loss": 0.4863,
+      "step": 121
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.899563263509725e-05,
+      "loss": 0.4434,
+      "step": 122
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.89757966579612e-05,
+      "loss": 0.443,
+      "step": 123
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8955777279528414e-05,
+      "loss": 0.4813,
+      "step": 124
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8935574908852272e-05,
+      "loss": 0.4609,
+      "step": 125
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8915189958725207e-05,
+      "loss": 0.4542,
+      "step": 126
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8894622845670282e-05,
+      "loss": 0.4285,
+      "step": 127
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8873873989932666e-05,
+      "loss": 0.4887,
+      "step": 128
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8852943815471058e-05,
+      "loss": 0.4553,
+      "step": 129
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8831832749949015e-05,
+      "loss": 0.4411,
+      "step": 130
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8810541224726217e-05,
+      "loss": 0.4653,
+      "step": 131
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.878906967484966e-05,
+      "loss": 0.4594,
+      "step": 132
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8767418539044753e-05,
+      "loss": 0.4602,
+      "step": 133
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8745588259706366e-05,
+      "loss": 0.4504,
+      "step": 134
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8723579282889784e-05,
+      "loss": 0.4721,
+      "step": 135
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8701392058301595e-05,
+      "loss": 0.4561,
+      "step": 136
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.86790270392905e-05,
+      "loss": 0.4376,
+      "step": 137
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.865648468283805e-05,
+      "loss": 0.4513,
+      "step": 138
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.863376544954931e-05,
+      "loss": 0.4416,
+      "step": 139
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8610869803643454e-05,
+      "loss": 0.4298,
+      "step": 140
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8587798212944255e-05,
+      "loss": 0.4596,
+      "step": 141
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.856455114887056e-05,
+      "loss": 0.4627,
+      "step": 142
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.854112908642663e-05,
+      "loss": 0.472,
+      "step": 143
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8517532504192456e-05,
+      "loss": 0.4671,
+      "step": 144
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.849376188431396e-05,
+      "loss": 0.4807,
+      "step": 145
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8469817712493148e-05,
+      "loss": 0.4428,
+      "step": 146
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8445700477978207e-05,
+      "loss": 0.4461,
+      "step": 147
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8421410673553475e-05,
+      "loss": 0.4507,
+      "step": 148
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8396948795529405e-05,
+      "loss": 0.4594,
+      "step": 149
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8372315343732395e-05,
+      "loss": 0.4423,
+      "step": 150
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8347510821494593e-05,
+      "loss": 0.4668,
+      "step": 151
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8322535735643604e-05,
+      "loss": 0.4286,
+      "step": 152
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8297390596492143e-05,
+      "loss": 0.4313,
+      "step": 153
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8272075917827597e-05,
+      "loss": 0.4507,
+      "step": 154
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.824659221690153e-05,
+      "loss": 0.4631,
+      "step": 155
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.822094001441913e-05,
+      "loss": 0.4839,
+      "step": 156
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8195119834528535e-05,
+      "loss": 0.4634,
+      "step": 157
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8169132204810157e-05,
+      "loss": 0.442,
+      "step": 158
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.814297765626589e-05,
+      "loss": 0.4671,
+      "step": 159
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8116656723308253e-05,
+      "loss": 0.4622,
+      "step": 160
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.4721,
+      "step": 161
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8063517858790517e-05,
+      "loss": 0.4728,
+      "step": 162
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8036701013009988e-05,
+      "loss": 0.4662,
+      "step": 163
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.800971995435305e-05,
+      "loss": 0.4916,
+      "step": 164
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7982575234120196e-05,
+      "loss": 0.4598,
+      "step": 165
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7955267406955997e-05,
+      "loss": 0.4434,
+      "step": 166
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.792779703083777e-05,
+      "loss": 0.4565,
+      "step": 167
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.790016466706417e-05,
+      "loss": 0.4633,
+      "step": 168
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.787237088024372e-05,
+      "loss": 0.4623,
+      "step": 169
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.78444162382833e-05,
+      "loss": 0.4559,
+      "step": 170
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.781630131237649e-05,
+      "loss": 0.4421,
+      "step": 171
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.778802667699196e-05,
+      "loss": 0.4357,
+      "step": 172
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7759592909861694e-05,
+      "loss": 0.4538,
+      "step": 173
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7731000591969182e-05,
+      "loss": 0.4342,
+      "step": 174
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7702250307537583e-05,
+      "loss": 0.4581,
+      "step": 175
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7673342644017744e-05,
+      "loss": 0.4485,
+      "step": 176
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.764427819207624e-05,
+      "loss": 0.4574,
+      "step": 177
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.761505754558327e-05,
+      "loss": 0.4474,
+      "step": 178
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.758568130160053e-05,
+      "loss": 0.4478,
+      "step": 179
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.755615006036904e-05,
+      "loss": 0.4543,
+      "step": 180
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7526464425296846e-05,
+      "loss": 0.4528,
+      "step": 181
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7496625002946702e-05,
+      "loss": 0.4313,
+      "step": 182
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.746663240302368e-05,
+      "loss": 0.4869,
+      "step": 183
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.743648723836271e-05,
+      "loss": 0.472,
+      "step": 184
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7406190124916064e-05,
+      "loss": 0.4585,
+      "step": 185
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.737574168174075e-05,
+      "loss": 0.4215,
+      "step": 186
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.734514253098589e-05,
+      "loss": 0.452,
+      "step": 187
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7314393297879982e-05,
+      "loss": 0.456,
+      "step": 188
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7283494610718153e-05,
+      "loss": 0.4557,
+      "step": 189
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7252447100849294e-05,
+      "loss": 0.4679,
+      "step": 190
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7221251402663176e-05,
+      "loss": 0.4364,
+      "step": 191
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7189908153577473e-05,
+      "loss": 0.4641,
+      "step": 192
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7158417994024766e-05,
+      "loss": 0.4387,
+      "step": 193
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7126781567439418e-05,
+      "loss": 0.4487,
+      "step": 194
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.709499952024447e-05,
+      "loss": 0.4381,
+      "step": 195
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7063072501838388e-05,
+      "loss": 0.4622,
+      "step": 196
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7031001164581828e-05,
+      "loss": 0.4528,
+      "step": 197
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6998786163784295e-05,
+      "loss": 0.4374,
+      "step": 198
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.696642815769075e-05,
+      "loss": 0.4468,
+      "step": 199
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6933927807468155e-05,
+      "loss": 0.4716,
+      "step": 200
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.690128577719199e-05,
+      "loss": 0.4599,
+      "step": 201
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6868502733832647e-05,
+      "loss": 0.4494,
+      "step": 202
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.683557934724183e-05,
+      "loss": 0.4539,
+      "step": 203
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.680251629013885e-05,
+      "loss": 0.445,
+      "step": 204
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6769314238096906e-05,
+      "loss": 0.4293,
+      "step": 205
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.673597386952924e-05,
+      "loss": 0.4578,
+      "step": 206
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.670249586567531e-05,
+      "loss": 0.434,
+      "step": 207
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6668880910586853e-05,
+      "loss": 0.4491,
+      "step": 208
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.663512969111392e-05,
+      "loss": 0.4461,
+      "step": 209
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6601242896890832e-05,
+      "loss": 0.4375,
+      "step": 210
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6567221220322082e-05,
+      "loss": 0.4456,
+      "step": 211
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6533065356568206e-05,
+      "loss": 0.4541,
+      "step": 212
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6498776003531575e-05,
+      "loss": 0.4385,
+      "step": 213
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6464353861842115e-05,
+      "loss": 0.4304,
+      "step": 214
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6429799634843012e-05,
+      "loss": 0.4517,
+      "step": 215
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6395114028576344e-05,
+      "loss": 0.4442,
+      "step": 216
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.636029775176862e-05,
+      "loss": 0.437,
+      "step": 217
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6325351515816353e-05,
+      "loss": 0.4193,
+      "step": 218
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.629027603477147e-05,
+      "loss": 0.4421,
+      "step": 219
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6255072025326763e-05,
+      "loss": 0.4511,
+      "step": 220
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.621974020680122e-05,
+      "loss": 0.4409,
+      "step": 221
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.618428130112533e-05,
+      "loss": 0.4239,
+      "step": 222
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6148696032826354e-05,
+      "loss": 0.4479,
+      "step": 223
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.611298512901349e-05,
+      "loss": 0.4601,
+      "step": 224
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6077149319363035e-05,
+      "loss": 0.435,
+      "step": 225
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6041189336103475e-05,
+      "loss": 0.4369,
+      "step": 226
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6005105914000508e-05,
+      "loss": 0.452,
+      "step": 227
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.596889979034205e-05,
+      "loss": 0.4772,
+      "step": 228
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5932571704923168e-05,
+      "loss": 0.4528,
+      "step": 229
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.589612240003095e-05,
+      "loss": 0.441,
+      "step": 230
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.585955262042934e-05,
+      "loss": 0.445,
+      "step": 231
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5822863113343934e-05,
+      "loss": 0.4712,
+      "step": 232
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5786054628446712e-05,
+      "loss": 0.4349,
+      "step": 233
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.57491279178407e-05,
+      "loss": 0.4509,
+      "step": 234
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5712083736044613e-05,
+      "loss": 0.435,
+      "step": 235
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5674922839977446e-05,
+      "loss": 0.4306,
+      "step": 236
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5637645988943008e-05,
+      "loss": 0.4595,
+      "step": 237
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.560025394461439e-05,
+      "loss": 0.4597,
+      "step": 238
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5562747471018415e-05,
+      "loss": 0.4572,
+      "step": 239
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.552512733452003e-05,
+      "loss": 0.376,
+      "step": 240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5487394303806632e-05,
+      "loss": 0.3759,
+      "step": 241
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.544954914987238e-05,
+      "loss": 0.3626,
+      "step": 242
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.541159264600242e-05,
+      "loss": 0.397,
+      "step": 243
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5373525567757124e-05,
+      "loss": 0.3663,
+      "step": 244
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5335348692956177e-05,
+      "loss": 0.3769,
+      "step": 245
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.529706280166276e-05,
+      "loss": 0.3705,
+      "step": 246
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5258668676167548e-05,
+      "loss": 0.3748,
+      "step": 247
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5220167100972763e-05,
+      "loss": 0.3654,
+      "step": 248
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.518155886277613e-05,
+      "loss": 0.3639,
+      "step": 249
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5142844750454807e-05,
+      "loss": 0.3645,
+      "step": 250
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5104025555049262e-05,
+      "loss": 0.3606,
+      "step": 251
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5065102069747117e-05,
+      "loss": 0.3607,
+      "step": 252
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.502607508986693e-05,
+      "loss": 0.35,
+      "step": 253
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.498694541284195e-05,
+      "loss": 0.3839,
+      "step": 254
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4947713838203835e-05,
+      "loss": 0.3718,
+      "step": 255
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4908381167566286e-05,
+      "loss": 0.3666,
+      "step": 256
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.48689482046087e-05,
+      "loss": 0.3587,
+      "step": 257
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4829415755059726e-05,
+      "loss": 0.3605,
+      "step": 258
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4789784626680819e-05,
+      "loss": 0.3575,
+      "step": 259
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.475005562924971e-05,
+      "loss": 0.3515,
+      "step": 260
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4710229574543893e-05,
+      "loss": 0.3763,
+      "step": 261
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.467030727632401e-05,
+      "loss": 0.3653,
+      "step": 262
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4630289550317234e-05,
+      "loss": 0.3454,
+      "step": 263
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4590177214200609e-05,
+      "loss": 0.3394,
+      "step": 264
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4549971087584329e-05,
+      "loss": 0.3479,
+      "step": 265
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4509671991995003e-05,
+      "loss": 0.3624,
+      "step": 266
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4469280750858854e-05,
+      "loss": 0.383,
+      "step": 267
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4428798189484914e-05,
+      "loss": 0.3767,
+      "step": 268
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4388225135048137e-05,
+      "loss": 0.3593,
+      "step": 269
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4347562416572525e-05,
+      "loss": 0.3576,
+      "step": 270
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.430681086491416e-05,
+      "loss": 0.3578,
+      "step": 271
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4265971312744252e-05,
+      "loss": 0.3413,
+      "step": 272
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4225044594532104e-05,
+      "loss": 0.3641,
+      "step": 273
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4184031546528077e-05,
+      "loss": 0.3536,
+      "step": 274
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4142933006746502e-05,
+      "loss": 0.3691,
+      "step": 275
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4101749814948544e-05,
+      "loss": 0.3735,
+      "step": 276
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4060482812625055e-05,
+      "loss": 0.3634,
+      "step": 277
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4019132842979375e-05,
+      "loss": 0.3649,
+      "step": 278
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3977700750910112e-05,
+      "loss": 0.3546,
+      "step": 279
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3936187382993862e-05,
+      "loss": 0.3672,
+      "step": 280
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3894593587467924e-05,
+      "loss": 0.3381,
+      "step": 281
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3852920214212966e-05,
+      "loss": 0.3708,
+      "step": 282
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3811168114735647e-05,
+      "loss": 0.3776,
+      "step": 283
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3769338142151245e-05,
+      "loss": 0.3588,
+      "step": 284
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3727431151166196e-05,
+      "loss": 0.3656,
+      "step": 285
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3685447998060651e-05,
+      "loss": 0.378,
+      "step": 286
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3643389540670963e-05,
+      "loss": 0.3652,
+      "step": 287
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3601256638372182e-05,
+      "loss": 0.3688,
+      "step": 288
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3559050152060465e-05,
+      "loss": 0.3577,
+      "step": 289
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3516770944135514e-05,
+      "loss": 0.3593,
+      "step": 290
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3474419878482935e-05,
+      "loss": 0.3425,
+      "step": 291
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3431997820456592e-05,
+      "loss": 0.38,
+      "step": 292
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3389505636860944e-05,
+      "loss": 0.3717,
+      "step": 293
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3346944195933294e-05,
+      "loss": 0.3476,
+      "step": 294
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.330431436732608e-05,
+      "loss": 0.3779,
+      "step": 295
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3261617022089103e-05,
+      "loss": 0.3492,
+      "step": 296
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3218853032651719e-05,
+      "loss": 0.3657,
+      "step": 297
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3176023272805008e-05,
+      "loss": 0.3524,
+      "step": 298
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.313312861768394e-05,
+      "loss": 0.3704,
+      "step": 299
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.3631,
+      "step": 300
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3047148128770664e-05,
+      "loss": 0.3806,
+      "step": 301
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3004064051806712e-05,
+      "loss": 0.3577,
+      "step": 302
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2960918593189005e-05,
+      "loss": 0.3629,
+      "step": 303
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2917712634503148e-05,
+      "loss": 0.3581,
+      "step": 304
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2874447058570927e-05,
+      "loss": 0.3801,
+      "step": 305
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2831122749432278e-05,
+      "loss": 0.3409,
+      "step": 306
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2787740592327232e-05,
+      "loss": 0.3563,
+      "step": 307
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2744301473677814e-05,
+      "loss": 0.3664,
+      "step": 308
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2700806281069942e-05,
+      "loss": 0.3702,
+      "step": 309
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2657255903235278e-05,
+      "loss": 0.3679,
+      "step": 310
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2613651230033085e-05,
+      "loss": 0.3626,
+      "step": 311
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2569993152432028e-05,
+      "loss": 0.3574,
+      "step": 312
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2526282562491991e-05,
+      "loss": 0.3684,
+      "step": 313
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2482520353345819e-05,
+      "loss": 0.3683,
+      "step": 314
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2438707419181097e-05,
+      "loss": 0.3704,
+      "step": 315
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2394844655221863e-05,
+      "loss": 0.3782,
+      "step": 316
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2350932957710322e-05,
+      "loss": 0.3704,
+      "step": 317
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2306973223888535e-05,
+      "loss": 0.3752,
+      "step": 318
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2262966351980075e-05,
+      "loss": 0.3592,
+      "step": 319
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2218913241171691e-05,
+      "loss": 0.3506,
+      "step": 320
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2174814791594913e-05,
+      "loss": 0.3413,
+      "step": 321
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2130671904307692e-05,
+      "loss": 0.3435,
+      "step": 322
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2086485481275943e-05,
+      "loss": 0.3685,
+      "step": 323
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2042256425355165e-05,
+      "loss": 0.3643,
+      "step": 324
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1997985640271956e-05,
+      "loss": 0.3758,
+      "step": 325
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1953674030605568e-05,
+      "loss": 0.3781,
+      "step": 326
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1909322501769407e-05,
+      "loss": 0.3549,
+      "step": 327
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.186493195999255e-05,
+      "loss": 0.363,
+      "step": 328
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1820503312301218e-05,
+      "loss": 0.3435,
+      "step": 329
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1776037466500245e-05,
+      "loss": 0.3626,
+      "step": 330
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1731535331154532e-05,
+      "loss": 0.3714,
+      "step": 331
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1686997815570473e-05,
+      "loss": 0.358,
+      "step": 332
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1642425829777391e-05,
+      "loss": 0.3592,
+      "step": 333
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1597820284508927e-05,
+      "loss": 0.3675,
+      "step": 334
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1553182091184439e-05,
+      "loss": 0.3545,
+      "step": 335
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1508512161890381e-05,
+      "loss": 0.3593,
+      "step": 336
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1463811409361667e-05,
+      "loss": 0.3663,
+      "step": 337
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1419080746963012e-05,
+      "loss": 0.3528,
+      "step": 338
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1374321088670277e-05,
+      "loss": 0.3664,
+      "step": 339
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1329533349051794e-05,
+      "loss": 0.3702,
+      "step": 340
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1284718443249676e-05,
+      "loss": 0.3712,
+      "step": 341
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1239877286961123e-05,
+      "loss": 0.3572,
+      "step": 342
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.11950107964197e-05,
+      "loss": 0.3646,
+      "step": 343
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1150119888376631e-05,
+      "loss": 0.3399,
+      "step": 344
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1105205480082052e-05,
+      "loss": 0.3683,
+      "step": 345
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.106026848926629e-05,
+      "loss": 0.356,
+      "step": 346
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1015309834121083e-05,
+      "loss": 0.3749,
+      "step": 347
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0970330433280838e-05,
+      "loss": 0.3572,
+      "step": 348
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0925331205803861e-05,
+      "loss": 0.3536,
+      "step": 349
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0880313071153568e-05,
+      "loss": 0.3611,
+      "step": 350
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0835276949179707e-05,
+      "loss": 0.3735,
+      "step": 351
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.079022376009955e-05,
+      "loss": 0.3728,
+      "step": 352
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0745154424479112e-05,
+      "loss": 0.3526,
+      "step": 353
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0700069863214317e-05,
+      "loss": 0.3574,
+      "step": 354
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0654970997512201e-05,
+      "loss": 0.3753,
+      "step": 355
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0609858748872073e-05,
+      "loss": 0.3666,
+      "step": 356
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.05647340390667e-05,
+      "loss": 0.346,
+      "step": 357
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0519597790123463e-05,
+      "loss": 0.3615,
+      "step": 358
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.047445092430552e-05,
+      "loss": 0.3601,
+      "step": 359
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0429294364092968e-05,
+      "loss": 0.3692,
+      "step": 360
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0384129032163976e-05,
+      "loss": 0.3886,
+      "step": 361
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0338955851375962e-05,
+      "loss": 0.3539,
+      "step": 362
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0293775744746705e-05,
+      "loss": 0.3746,
+      "step": 363
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0248589635435505e-05,
+      "loss": 0.3851,
+      "step": 364
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0203398446724306e-05,
+      "loss": 0.3546,
+      "step": 365
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0158203101998854e-05,
+      "loss": 0.3599,
+      "step": 366
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.01130045247298e-05,
+      "loss": 0.3639,
+      "step": 367
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0067803638453847e-05,
+      "loss": 0.3592,
+      "step": 368
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0022601366754889e-05,
+      "loss": 0.3565,
+      "step": 369
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.977398633245116e-06,
+      "loss": 0.3535,
+      "step": 370
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.932196361546156e-06,
+      "loss": 0.354,
+      "step": 371
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.886995475270205e-06,
+      "loss": 0.3416,
+      "step": 372
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.84179689800115e-06,
+      "loss": 0.3462,
+      "step": 373
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.796601553275697e-06,
+      "loss": 0.3709,
+      "step": 374
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.751410364564499e-06,
+      "loss": 0.3574,
+      "step": 375
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.706224255253297e-06,
+      "loss": 0.3694,
+      "step": 376
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.661044148624038e-06,
+      "loss": 0.354,
+      "step": 377
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.615870967836026e-06,
+      "loss": 0.3834,
+      "step": 378
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.570705635907038e-06,
+      "loss": 0.3805,
+      "step": 379
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.525549075694484e-06,
+      "loss": 0.3694,
+      "step": 380
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.48040220987654e-06,
+      "loss": 0.3533,
+      "step": 381
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.435265960933304e-06,
+      "loss": 0.3654,
+      "step": 382
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.39014125112793e-06,
+      "loss": 0.3586,
+      "step": 383
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.3450290024878e-06,
+      "loss": 0.3707,
+      "step": 384
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.299930136785685e-06,
+      "loss": 0.3409,
+      "step": 385
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.25484557552089e-06,
+      "loss": 0.3458,
+      "step": 386
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.209776239900453e-06,
+      "loss": 0.3624,
+      "step": 387
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.164723050820298e-06,
+      "loss": 0.3684,
+      "step": 388
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.119686928846437e-06,
+      "loss": 0.3856,
+      "step": 389
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.074668794196142e-06,
+      "loss": 0.3841,
+      "step": 390
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.029669566719165e-06,
+      "loss": 0.3658,
+      "step": 391
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.98469016587892e-06,
+      "loss": 0.3515,
+      "step": 392
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.939731510733711e-06,
+      "loss": 0.3672,
+      "step": 393
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.894794519917947e-06,
+      "loss": 0.386,
+      "step": 394
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.849880111623374e-06,
+      "loss": 0.3662,
+      "step": 395
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.804989203580303e-06,
+      "loss": 0.3821,
+      "step": 396
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.76012271303888e-06,
+      "loss": 0.3603,
+      "step": 397
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.715281556750327e-06,
+      "loss": 0.3587,
+      "step": 398
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.670466650948208e-06,
+      "loss": 0.3516,
+      "step": 399
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.625678911329727e-06,
+      "loss": 0.3782,
+      "step": 400
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.580919253036991e-06,
+      "loss": 0.3443,
+      "step": 401
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.536188590638334e-06,
+      "loss": 0.3498,
+      "step": 402
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.491487838109622e-06,
+      "loss": 0.3547,
+      "step": 403
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.446817908815566e-06,
+      "loss": 0.3432,
+      "step": 404
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.402179715491078e-06,
+      "loss": 0.3499,
+      "step": 405
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.357574170222612e-06,
+      "loss": 0.3522,
+      "step": 406
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.313002184429529e-06,
+      "loss": 0.3561,
+      "step": 407
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.268464668845471e-06,
+      "loss": 0.345,
+      "step": 408
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.223962533499757e-06,
+      "loss": 0.358,
+      "step": 409
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.179496687698785e-06,
+      "loss": 0.3514,
+      "step": 410
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.135068040007452e-06,
+      "loss": 0.3495,
+      "step": 411
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.090677498230598e-06,
+      "loss": 0.3693,
+      "step": 412
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.046325969394437e-06,
+      "loss": 0.3638,
+      "step": 413
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.002014359728046e-06,
+      "loss": 0.363,
+      "step": 414
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.957743574644837e-06,
+      "loss": 0.3701,
+      "step": 415
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.913514518724059e-06,
+      "loss": 0.3653,
+      "step": 416
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.869328095692313e-06,
+      "loss": 0.3685,
+      "step": 417
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.825185208405089e-06,
+      "loss": 0.3587,
+      "step": 418
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.781086758828314e-06,
+      "loss": 0.3902,
+      "step": 419
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.73703364801993e-06,
+      "loss": 0.3687,
+      "step": 420
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.69302677611147e-06,
+      "loss": 0.3711,
+      "step": 421
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.649067042289681e-06,
+      "loss": 0.3389,
+      "step": 422
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.6051553447781415e-06,
+      "loss": 0.3502,
+      "step": 423
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.561292580818906e-06,
+      "loss": 0.3586,
+      "step": 424
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.517479646654184e-06,
+      "loss": 0.3741,
+      "step": 425
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.47371743750801e-06,
+      "loss": 0.3612,
+      "step": 426
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.430006847567972e-06,
+      "loss": 0.3683,
+      "step": 427
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.386348769966918e-06,
+      "loss": 0.3602,
+      "step": 428
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.342744096764727e-06,
+      "loss": 0.3432,
+      "step": 429
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.299193718930062e-06,
+      "loss": 0.3879,
+      "step": 430
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.255698526322188e-06,
+      "loss": 0.3776,
+      "step": 431
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.2122594076727705e-06,
+      "loss": 0.3681,
+      "step": 432
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.1688772505677225e-06,
+      "loss": 0.3502,
+      "step": 433
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.125552941429077e-06,
+      "loss": 0.3609,
+      "step": 434
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.082287365496852e-06,
+      "loss": 0.3782,
+      "step": 435
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.0390814068109965e-06,
+      "loss": 0.3778,
+      "step": 436
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.995935948193294e-06,
+      "loss": 0.3512,
+      "step": 437
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.9528518712293405e-06,
+      "loss": 0.371,
+      "step": 438
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.3671,
+      "step": 439
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.866871382316063e-06,
+      "loss": 0.3607,
+      "step": 440
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.823976727194994e-06,
+      "loss": 0.3655,
+      "step": 441
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.781146967348283e-06,
+      "loss": 0.3604,
+      "step": 442
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.738382977910898e-06,
+      "loss": 0.3622,
+      "step": 443
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.695685632673919e-06,
+      "loss": 0.3552,
+      "step": 444
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.653055804066712e-06,
+      "loss": 0.3643,
+      "step": 445
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.6104943631390596e-06,
+      "loss": 0.351,
+      "step": 446
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.568002179543409e-06,
+      "loss": 0.3676,
+      "step": 447
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.525580121517069e-06,
+      "loss": 0.3563,
+      "step": 448
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.48322905586449e-06,
+      "loss": 0.3502,
+      "step": 449
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.440949847939538e-06,
+      "loss": 0.3845,
+      "step": 450
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.39874336162782e-06,
+      "loss": 0.3567,
+      "step": 451
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.356610459329038e-06,
+      "loss": 0.3552,
+      "step": 452
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.314552001939351e-06,
+      "loss": 0.353,
+      "step": 453
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.272568848833809e-06,
+      "loss": 0.3603,
+      "step": 454
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.230661857848759e-06,
+      "loss": 0.3501,
+      "step": 455
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.188831885264357e-06,
+      "loss": 0.3576,
+      "step": 456
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.147079785787038e-06,
+      "loss": 0.3458,
+      "step": 457
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.105406412532078e-06,
+      "loss": 0.3639,
+      "step": 458
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.06381261700614e-06,
+      "loss": 0.3604,
+      "step": 459
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.022299249089889e-06,
+      "loss": 0.358,
+      "step": 460
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.980867157020624e-06,
+      "loss": 0.3592,
+      "step": 461
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.93951718737495e-06,
+      "loss": 0.3664,
+      "step": 462
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.8982501850514614e-06,
+      "loss": 0.351,
+      "step": 463
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.857066993253501e-06,
+      "loss": 0.3587,
+      "step": 464
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.815968453471923e-06,
+      "loss": 0.3763,
+      "step": 465
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.7749554054679015e-06,
+      "loss": 0.3431,
+      "step": 466
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.7340286872557515e-06,
+      "loss": 0.3664,
+      "step": 467
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.693189135085839e-06,
+      "loss": 0.3623,
+      "step": 468
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.652437583427478e-06,
+      "loss": 0.3609,
+      "step": 469
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.6117748649518665e-06,
+      "loss": 0.3651,
+      "step": 470
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5712018105150914e-06,
+      "loss": 0.3439,
+      "step": 471
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.530719249141148e-06,
+      "loss": 0.3526,
+      "step": 472
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.490328008005002e-06,
+      "loss": 0.3275,
+      "step": 473
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.450028912415672e-06,
+      "loss": 0.3615,
+      "step": 474
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.409822785799393e-06,
+      "loss": 0.3678,
+      "step": 475
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.369710449682767e-06,
+      "loss": 0.3561,
+      "step": 476
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.329692723675994e-06,
+      "loss": 0.37,
+      "step": 477
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.289770425456109e-06,
+      "loss": 0.3358,
+      "step": 478
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.249944370750293e-06,
+      "loss": 0.2953,
+      "step": 479
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.210215373319183e-06,
+      "loss": 0.282,
+      "step": 480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.170584244940275e-06,
+      "loss": 0.2803,
+      "step": 481
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.131051795391302e-06,
+      "loss": 0.3133,
+      "step": 482
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.091618832433716e-06,
+      "loss": 0.2831,
+      "step": 483
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.0522861617961694e-06,
+      "loss": 0.2965,
+      "step": 484
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.0130545871580504e-06,
+      "loss": 0.2828,
+      "step": 485
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.973924910133071e-06,
+      "loss": 0.277,
+      "step": 486
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.934897930252887e-06,
+      "loss": 0.2744,
+      "step": 487
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.895974444950743e-06,
+      "loss": 0.2799,
+      "step": 488
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.857155249545197e-06,
+      "loss": 0.2943,
+      "step": 489
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.8184411372238724e-06,
+      "loss": 0.2936,
+      "step": 490
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.779832899027243e-06,
+      "loss": 0.2746,
+      "step": 491
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.7413313238324556e-06,
+      "loss": 0.2795,
+      "step": 492
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.702937198337241e-06,
+      "loss": 0.2914,
+      "step": 493
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.66465130704382e-06,
+      "loss": 0.2793,
+      "step": 494
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.626474432242879e-06,
+      "loss": 0.2842,
+      "step": 495
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.58840735399758e-06,
+      "loss": 0.2617,
+      "step": 496
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.550450850127626e-06,
+      "loss": 0.2662,
+      "step": 497
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.512605696193371e-06,
+      "loss": 0.2748,
+      "step": 498
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.474872665479974e-06,
+      "loss": 0.2757,
+      "step": 499
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.437252528981586e-06,
+      "loss": 0.2774,
+      "step": 500
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.3997460553856095e-06,
+      "loss": 0.2918,
+      "step": 501
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.3623540110569935e-06,
+      "loss": 0.2629,
+      "step": 502
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.3250771600225536e-06,
+      "loss": 0.2864,
+      "step": 503
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.2879162639553925e-06,
+      "loss": 0.2888,
+      "step": 504
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.250872082159305e-06,
+      "loss": 0.2739,
+      "step": 505
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.213945371553292e-06,
+      "loss": 0.2943,
+      "step": 506
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.177136886656067e-06,
+      "loss": 0.2941,
+      "step": 507
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.140447379570663e-06,
+      "loss": 0.2793,
+      "step": 508
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.103877599969056e-06,
+      "loss": 0.2721,
+      "step": 509
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.067428295076833e-06,
+      "loss": 0.2887,
+      "step": 510
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.0311002096579486e-06,
+      "loss": 0.2767,
+      "step": 511
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.9948940859994964e-06,
+      "loss": 0.2794,
+      "step": 512
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.958810663896531e-06,
+      "loss": 0.2671,
+      "step": 513
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.922850680636968e-06,
+      "loss": 0.2853,
+      "step": 514
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.8870148709865115e-06,
+      "loss": 0.2874,
+      "step": 515
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.851303967173647e-06,
+      "loss": 0.285,
+      "step": 516
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.815718698874672e-06,
+      "loss": 0.2774,
+      "step": 517
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.780259793198784e-06,
+      "loss": 0.2798,
+      "step": 518
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.744927974673237e-06,
+      "loss": 0.2784,
+      "step": 519
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.709723965228531e-06,
+      "loss": 0.2739,
+      "step": 520
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6746484841836516e-06,
+      "loss": 0.2825,
+      "step": 521
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6397022482313804e-06,
+      "loss": 0.2929,
+      "step": 522
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.6048859714236597e-06,
+      "loss": 0.2661,
+      "step": 523
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.5702003651569883e-06,
+      "loss": 0.2726,
+      "step": 524
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.5356461381578865e-06,
+      "loss": 0.2776,
+      "step": 525
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.501223996468426e-06,
+      "loss": 0.2857,
+      "step": 526
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.466934643431795e-06,
+      "loss": 0.2654,
+      "step": 527
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.432778779677921e-06,
+      "loss": 0.2733,
+      "step": 528
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.3987571031091735e-06,
+      "loss": 0.2708,
+      "step": 529
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.36487030888608e-06,
+      "loss": 0.2654,
+      "step": 530
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3311190894131495e-06,
+      "loss": 0.2674,
+      "step": 531
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2975041343246937e-06,
+      "loss": 0.2899,
+      "step": 532
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.264026130470762e-06,
+      "loss": 0.2919,
+      "step": 533
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.230685761903094e-06,
+      "loss": 0.2797,
+      "step": 534
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1974837098611487e-06,
+      "loss": 0.2717,
+      "step": 535
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1644206527581734e-06,
+      "loss": 0.2965,
+      "step": 536
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.1314972661673572e-06,
+      "loss": 0.2806,
+      "step": 537
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.0987142228080137e-06,
+      "loss": 0.2781,
+      "step": 538
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0660721925318483e-06,
+      "loss": 0.2628,
+      "step": 539
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0335718423092553e-06,
+      "loss": 0.2812,
+      "step": 540
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0012138362157062e-06,
+      "loss": 0.2826,
+      "step": 541
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9689988354181742e-06,
+      "loss": 0.2956,
+      "step": 542
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9369274981616137e-06,
+      "loss": 0.276,
+      "step": 543
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.905000479755531e-06,
+      "loss": 0.2887,
+      "step": 544
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8732184325605815e-06,
+      "loss": 0.2927,
+      "step": 545
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8415820059752397e-06,
+      "loss": 0.3025,
+      "step": 546
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8100918464225304e-06,
+      "loss": 0.2713,
+      "step": 547
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.7787485973368288e-06,
+      "loss": 0.2926,
+      "step": 548
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.7475528991507106e-06,
+      "loss": 0.2858,
+      "step": 549
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.7165053892818495e-06,
+      "loss": 0.2805,
+      "step": 550
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.685606702120019e-06,
+      "loss": 0.2966,
+      "step": 551
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.654857469014113e-06,
+      "loss": 0.2799,
+      "step": 552
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.624258318259253e-06,
+      "loss": 0.2879,
+      "step": 553
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5938098750839414e-06,
+      "loss": 0.2838,
+      "step": 554
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.563512761637291e-06,
+      "loss": 0.287,
+      "step": 555
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.5333675969763215e-06,
+      "loss": 0.2933,
+      "step": 556
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.5033749970533015e-06,
+      "loss": 0.2679,
+      "step": 557
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.4735355747031566e-06,
+      "loss": 0.2697,
+      "step": 558
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.443849939630959e-06,
+      "loss": 0.2877,
+      "step": 559
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4143186983994715e-06,
+      "loss": 0.2805,
+      "step": 560
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.384942454416734e-06,
+      "loss": 0.2653,
+      "step": 561
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3557218079237608e-06,
+      "loss": 0.2629,
+      "step": 562
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3266573559822568e-06,
+      "loss": 0.2791,
+      "step": 563
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.2977496924624223e-06,
+      "loss": 0.2759,
+      "step": 564
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.26899940803082e-06,
+      "loss": 0.299,
+      "step": 565
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.240407090138309e-06,
+      "loss": 0.2782,
+      "step": 566
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.211973323008041e-06,
+      "loss": 0.2876,
+      "step": 567
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.183698687623511e-06,
+      "loss": 0.2667,
+      "step": 568
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.155583761716703e-06,
+      "loss": 0.2801,
+      "step": 569
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1276291197562772e-06,
+      "loss": 0.2823,
+      "step": 570
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0998353329358355e-06,
+      "loss": 0.2687,
+      "step": 571
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.072202969162234e-06,
+      "loss": 0.2793,
+      "step": 572
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0447325930440043e-06,
+      "loss": 0.2905,
+      "step": 573
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0174247658798054e-06,
+      "loss": 0.2711,
+      "step": 574
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.990280045646954e-06,
+      "loss": 0.2812,
+      "step": 575
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9632989869900145e-06,
+      "loss": 0.2889,
+      "step": 576
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.936482141209486e-06,
+      "loss": 0.2894,
+      "step": 577
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.287,
+      "step": 578
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8833432766917514e-06,
+      "loss": 0.265,
+      "step": 579
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8570223437341118e-06,
+      "loss": 0.267,
+      "step": 580
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8308677951898435e-06,
+      "loss": 0.2806,
+      "step": 581
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.8048801654714687e-06,
+      "loss": 0.2855,
+      "step": 582
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7790599855808732e-06,
+      "loss": 0.2801,
+      "step": 583
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7534077830984697e-06,
+      "loss": 0.3075,
+      "step": 584
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7279240821724063e-06,
+      "loss": 0.2642,
+      "step": 585
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7026094035078589e-06,
+      "loss": 0.2687,
+      "step": 586
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6774642643563955e-06,
+      "loss": 0.292,
+      "step": 587
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6524891785054097e-06,
+      "loss": 0.2845,
+      "step": 588
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6276846562676085e-06,
+      "loss": 0.2811,
+      "step": 589
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.603051204470597e-06,
+      "loss": 0.2693,
+      "step": 590
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5785893264465257e-06,
+      "loss": 0.2862,
+      "step": 591
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5542995220217961e-06,
+      "loss": 0.2855,
+      "step": 592
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.530182287506855e-06,
+      "loss": 0.2739,
+      "step": 593
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.506238115686044e-06,
+      "loss": 0.2747,
+      "step": 594
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4824674958075436e-06,
+      "loss": 0.2855,
+      "step": 595
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.458870913573368e-06,
+      "loss": 0.2549,
+      "step": 596
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4354488511294418e-06,
+      "loss": 0.2839,
+      "step": 597
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4122017870557458e-06,
+      "loss": 0.2761,
+      "step": 598
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3891301963565473e-06,
+      "loss": 0.2748,
+      "step": 599
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3662345504506903e-06,
+      "loss": 0.2749,
+      "step": 600
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.343515317161952e-06,
+      "loss": 0.2668,
+      "step": 601
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3209729607095022e-06,
+      "loss": 0.269,
+      "step": 602
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.2986079416984088e-06,
+      "loss": 0.262,
+      "step": 603
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2764207171102206e-06,
+      "loss": 0.2749,
+      "step": 604
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2544117402936373e-06,
+      "loss": 0.2641,
+      "step": 605
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.232581460955249e-06,
+      "loss": 0.2747,
+      "step": 606
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2109303251503434e-06,
+      "loss": 0.2792,
+      "step": 607
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.189458775273784e-06,
+      "loss": 0.294,
+      "step": 608
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1681672500509866e-06,
+      "loss": 0.2683,
+      "step": 609
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.147056184528943e-06,
+      "loss": 0.2803,
+      "step": 610
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1261260100673355e-06,
+      "loss": 0.2867,
+      "step": 611
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1053771543297198e-06,
+      "loss": 0.2863,
+      "step": 612
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.0848100412747954e-06,
+      "loss": 0.2759,
+      "step": 613
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0644250911477306e-06,
+      "loss": 0.269,
+      "step": 614
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0442227204715872e-06,
+      "loss": 0.2931,
+      "step": 615
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0242033420388008e-06,
+      "loss": 0.2823,
+      "step": 616
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0043673649027519e-06,
+      "loss": 0.2825,
+      "step": 617
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.8471519436941e-07,
+      "loss": 0.2799,
+      "step": 618
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.652472319890372e-07,
+      "loss": 0.2811,
+      "step": 619
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.459638755480038e-07,
+      "loss": 0.271,
+      "step": 620
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.268655190606501e-07,
+      "loss": 0.2781,
+      "step": 621
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.079525527612321e-07,
+      "loss": 0.275,
+      "step": 622
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.892253630959502e-07,
+      "loss": 0.2699,
+      "step": 623
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.706843327150605e-07,
+      "loss": 0.2765,
+      "step": 624
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.523298404650504e-07,
+      "loss": 0.2713,
+      "step": 625
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.34162261380892e-07,
+      "loss": 0.286,
+      "step": 626
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.161819666783888e-07,
+      "loss": 0.2642,
+      "step": 627
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.983893237465878e-07,
+      "loss": 0.2822,
+      "step": 628
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.807846961402699e-07,
+      "loss": 0.2743,
+      "step": 629
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.633684435725208e-07,
+      "loss": 0.2573,
+      "step": 630
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.461409219073857e-07,
+      "loss": 0.2813,
+      "step": 631
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.291024831525961e-07,
+      "loss": 0.2854,
+      "step": 632
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.122534754523768e-07,
+      "loss": 0.2888,
+      "step": 633
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.955942430803298e-07,
+      "loss": 0.27,
+      "step": 634
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.791251264324039e-07,
+      "loss": 0.2792,
+      "step": 635
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.628464620199404e-07,
+      "loss": 0.2774,
+      "step": 636
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.467585824627886e-07,
+      "loss": 0.2874,
+      "step": 637
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.30861816482522e-07,
+      "loss": 0.2792,
+      "step": 638
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.151564888957084e-07,
+      "loss": 0.2688,
+      "step": 639
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.996429206072874e-07,
+      "loss": 0.2653,
+      "step": 640
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.843214286039956e-07,
+      "loss": 0.2846,
+      "step": 641
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.691923259479093e-07,
+      "loss": 0.275,
+      "step": 642
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.542559217700339e-07,
+      "loss": 0.2796,
+      "step": 643
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.395125212639895e-07,
+      "loss": 0.2788,
+      "step": 644
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.249624256797803e-07,
+      "loss": 0.2688,
+      "step": 645
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.106059323176371e-07,
+      "loss": 0.2817,
+      "step": 646
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.964433345219354e-07,
+      "loss": 0.2784,
+      "step": 647
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.824749216752134e-07,
+      "loss": 0.2789,
+      "step": 648
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.6870097919224923e-07,
+      "loss": 0.2691,
+      "step": 649
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.551217885142378e-07,
+      "loss": 0.2759,
+      "step": 650
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.417376271030327e-07,
+      "loss": 0.2658,
+      "step": 651
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.285487684354772e-07,
+      "loss": 0.2678,
+      "step": 652
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.1555548199782357e-07,
+      "loss": 0.2867,
+      "step": 653
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.0275803328021946e-07,
+      "loss": 0.2882,
+      "step": 654
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.9015668377128446e-07,
+      "loss": 0.2879,
+      "step": 655
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.777516909527701e-07,
+      "loss": 0.2676,
+      "step": 656
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.6554330829429716e-07,
+      "loss": 0.2806,
+      "step": 657
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.5353178524817566e-07,
+      "loss": 0.2793,
+      "step": 658
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.417173672443075e-07,
+      "loss": 0.28,
+      "step": 659
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.30100295685174e-07,
+      "loss": 0.281,
+      "step": 660
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.1868080794090316e-07,
+      "loss": 0.288,
+      "step": 661
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.0745913734441357e-07,
+      "loss": 0.2726,
+      "step": 662
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.9643551318665917e-07,
+      "loss": 0.2707,
+      "step": 663
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.8561016071192884e-07,
+      "loss": 0.2706,
+      "step": 664
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.7498330111325635e-07,
+      "loss": 0.2862,
+      "step": 665
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.6455515152789435e-07,
+      "loss": 0.2803,
+      "step": 666
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.5432592503288e-07,
+      "loss": 0.288,
+      "step": 667
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.442958306406795e-07,
+      "loss": 0.2819,
+      "step": 668
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.3446507329492274e-07,
+      "loss": 0.2679,
+      "step": 669
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2483385386620317e-07,
+      "loss": 0.2654,
+      "step": 670
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.1540236914799116e-07,
+      "loss": 0.2812,
+      "step": 671
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0617081185259512e-07,
+      "loss": 0.2898,
+      "step": 672
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.9713937060723887e-07,
+      "loss": 0.2848,
+      "step": 673
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.8830822995019593e-07,
+      "loss": 0.29,
+      "step": 674
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.7967757032702481e-07,
+      "loss": 0.2836,
+      "step": 675
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.7124756808688525e-07,
+      "loss": 0.2878,
+      "step": 676
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.630183954789233e-07,
+      "loss": 0.2695,
+      "step": 677
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.5499022064876412e-07,
+      "loss": 0.2855,
+      "step": 678
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4716320763507152e-07,
+      "loss": 0.2836,
+      "step": 679
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.3953751636619162e-07,
+      "loss": 0.2726,
+      "step": 680
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.3211330265689104e-07,
+      "loss": 0.2808,
+      "step": 681
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2489071820517394e-07,
+      "loss": 0.2708,
+      "step": 682
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1786991058917785e-07,
+      "loss": 0.2905,
+      "step": 683
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1105102326415929e-07,
+      "loss": 0.2977,
+      "step": 684
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.0443419555956402e-07,
+      "loss": 0.2726,
+      "step": 685
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.801956267618262e-08,
+      "loss": 0.2702,
+      "step": 686
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.180725568338045e-08,
+      "loss": 0.2987,
+      "step": 687
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.579740151642534e-08,
+      "loss": 0.2706,
+      "step": 688
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.999012297389419e-08,
+      "loss": 0.2755,
+      "step": 689
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.438553871516152e-08,
+      "loss": 0.28,
+      "step": 690
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.898376325797596e-08,
+      "loss": 0.2785,
+      "step": 691
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.378490697611761e-08,
+      "loss": 0.2775,
+      "step": 692
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.878907609714879e-08,
+      "loss": 0.2705,
+      "step": 693
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.399637270023683e-08,
+      "loss": 0.2546,
+      "step": 694
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.940689471407356e-08,
+      "loss": 0.2658,
+      "step": 695
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.502073591487244e-08,
+      "loss": 0.2786,
+      "step": 696
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 4.083798592444899e-08,
+      "loss": 0.2753,
+      "step": 697
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.68587302083967e-08,
+      "loss": 0.2914,
+      "step": 698
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.308305007433399e-08,
+      "loss": 0.2827,
+      "step": 699
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.9511022670246635e-08,
+      "loss": 0.2731,
+      "step": 700
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.6142720982911264e-08,
+      "loss": 0.2901,
+      "step": 701
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.2978213836400974e-08,
+      "loss": 0.2711,
+      "step": 702
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.0017565890683154e-08,
+      "loss": 0.2774,
+      "step": 703
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.726083764029607e-08,
+      "loss": 0.2903,
+      "step": 704
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.4708085413113194e-08,
+      "loss": 0.2848,
+      "step": 705
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.2359361369190804e-08,
+      "loss": 0.2837,
+      "step": 706
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0214713499706596e-08,
+      "loss": 0.2665,
+      "step": 707
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.274185625971598e-09,
+      "loss": 0.2779,
+      "step": 708
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 6.5378173985441994e-09,
+      "loss": 0.2707,
+      "step": 709
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.0056442964119265e-09,
+      "loss": 0.2554,
+      "step": 710
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.6776976262697937e-09,
+      "loss": 0.2767,
+      "step": 711
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.5540045218819256e-09,
+      "loss": 0.2892,
+      "step": 712
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.6345879435231138e-09,
+      "loss": 0.2843,
+      "step": 713
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 9.194666775158567e-10,
+      "loss": 0.2735,
+      "step": 714
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 4.086553358395584e-10,
+      "loss": 0.2775,
+      "step": 715
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.0216435583743079e-10,
+      "loss": 0.2786,
+      "step": 716
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0,
+      "loss": 0.2579,
+      "step": 717
+    },
+    {
+      "epoch": 3.0,
+      "step": 717,
+      "total_flos": 1.9663471898819297e+18,
+      "train_loss": 0.3724504075788554,
+      "train_runtime": 6685.3329,
+      "train_samples_per_second": 54.728,
+      "train_steps_per_second": 0.107
+    }
+  ],
+  "max_steps": 717,
+  "num_train_epochs": 3,
+  "total_flos": 1.9663471898819297e+18,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:142478985dadb6f14a40dcb1885e8cd0d0a543b018a64abaa6a968bfe45980f1
+size 4155