Model save
Browse files- README.md +58 -0
- all_results.json +9 -0
- config.json +30 -0
- generation_config.json +6 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- special_tokens_map.json +34 -0
- tokenizer.json +0 -0
- tokenizer_config.json +155 -0
- train_results.json +9 -0
- trainer_state.json +2997 -0
- training_args.bin +3 -0
- vocab.json +0 -0
README.md
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: HuggingFaceTB/SmolLM2-1.7B
|
| 3 |
+
library_name: transformers
|
| 4 |
+
model_name: SmolLM2-MagPiePro
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- sft
|
| 9 |
+
licence: license
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Model Card for SmolLM2-MagPiePro
|
| 13 |
+
|
| 14 |
+
This model is a fine-tuned version of [HuggingFaceTB/SmolLM2-1.7B](https://huggingface.co/HuggingFaceTB/SmolLM2-1.7B).
|
| 15 |
+
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 16 |
+
|
| 17 |
+
## Quick start
|
| 18 |
+
|
| 19 |
+
```python
|
| 20 |
+
from transformers import pipeline
|
| 21 |
+
|
| 22 |
+
question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
|
| 23 |
+
generator = pipeline("text-generation", model="HuggingFaceTB/SmolLM2-MagPiePro", device="cuda")
|
| 24 |
+
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
|
| 25 |
+
print(output["generated_text"])
|
| 26 |
+
```
|
| 27 |
+
|
| 28 |
+
## Training procedure
|
| 29 |
+
|
| 30 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/loubnabnl/huggingface/runs/iiqtfsij)
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
This model was trained with SFT.
|
| 34 |
+
|
| 35 |
+
### Framework versions
|
| 36 |
+
|
| 37 |
+
- TRL: 0.14.0.dev0
|
| 38 |
+
- Transformers: 4.48.1
|
| 39 |
+
- Pytorch: 2.5.1+cu121
|
| 40 |
+
- Datasets: 3.2.0
|
| 41 |
+
- Tokenizers: 0.21.0
|
| 42 |
+
|
| 43 |
+
## Citations
|
| 44 |
+
|
| 45 |
+
|
| 46 |
+
|
| 47 |
+
Cite TRL as:
|
| 48 |
+
|
| 49 |
+
```bibtex
|
| 50 |
+
@misc{vonwerra2022trl,
|
| 51 |
+
title = {{TRL: Transformer Reinforcement Learning}},
|
| 52 |
+
author = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
|
| 53 |
+
year = 2020,
|
| 54 |
+
journal = {GitHub repository},
|
| 55 |
+
publisher = {GitHub},
|
| 56 |
+
howpublished = {\url{https://github.com/huggingface/trl}}
|
| 57 |
+
}
|
| 58 |
+
```
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9997629770087698,
|
| 3 |
+
"total_flos": 141388749045760.0,
|
| 4 |
+
"train_loss": 0.6815844999777304,
|
| 5 |
+
"train_runtime": 3835.2311,
|
| 6 |
+
"train_samples": 270000,
|
| 7 |
+
"train_samples_per_second": 70.4,
|
| 8 |
+
"train_steps_per_second": 0.55
|
| 9 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "HuggingFaceTB/SmolLM2-1.7B",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"LlamaForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"bos_token_id": 0,
|
| 9 |
+
"eos_token_id": 0,
|
| 10 |
+
"head_dim": 64,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 2048,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 8192,
|
| 15 |
+
"max_position_embeddings": 8192,
|
| 16 |
+
"mlp_bias": false,
|
| 17 |
+
"model_type": "llama",
|
| 18 |
+
"num_attention_heads": 32,
|
| 19 |
+
"num_hidden_layers": 24,
|
| 20 |
+
"num_key_value_heads": 32,
|
| 21 |
+
"pretraining_tp": 1,
|
| 22 |
+
"rms_norm_eps": 1e-05,
|
| 23 |
+
"rope_scaling": null,
|
| 24 |
+
"rope_theta": 130000,
|
| 25 |
+
"tie_word_embeddings": true,
|
| 26 |
+
"torch_dtype": "bfloat16",
|
| 27 |
+
"transformers_version": "4.48.1",
|
| 28 |
+
"use_cache": false,
|
| 29 |
+
"vocab_size": 49152
|
| 30 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"eos_token_id": 0,
|
| 5 |
+
"transformers_version": "4.48.1"
|
| 6 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e7986fd8ea03d19cf5512d488ad1926d9a5c57f56b296920cc4df2a7c82b38d4
|
| 3 |
+
size 3422777952
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|im_start|>",
|
| 4 |
+
"<|im_end|>"
|
| 5 |
+
],
|
| 6 |
+
"bos_token": {
|
| 7 |
+
"content": "<|im_start|>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": false,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false
|
| 12 |
+
},
|
| 13 |
+
"eos_token": {
|
| 14 |
+
"content": "<|im_end|>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false
|
| 19 |
+
},
|
| 20 |
+
"pad_token": {
|
| 21 |
+
"content": "<|im_end|>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": false,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false
|
| 26 |
+
},
|
| 27 |
+
"unk_token": {
|
| 28 |
+
"content": "<|endoftext|>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false
|
| 33 |
+
}
|
| 34 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,155 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"0": {
|
| 5 |
+
"content": "<|endoftext|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": false,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"1": {
|
| 13 |
+
"content": "<|im_start|>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": false,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"2": {
|
| 21 |
+
"content": "<|im_end|>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": false,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
},
|
| 28 |
+
"3": {
|
| 29 |
+
"content": "<repo_name>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": false,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false,
|
| 34 |
+
"special": true
|
| 35 |
+
},
|
| 36 |
+
"4": {
|
| 37 |
+
"content": "<reponame>",
|
| 38 |
+
"lstrip": false,
|
| 39 |
+
"normalized": false,
|
| 40 |
+
"rstrip": false,
|
| 41 |
+
"single_word": false,
|
| 42 |
+
"special": true
|
| 43 |
+
},
|
| 44 |
+
"5": {
|
| 45 |
+
"content": "<file_sep>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false,
|
| 50 |
+
"special": true
|
| 51 |
+
},
|
| 52 |
+
"6": {
|
| 53 |
+
"content": "<filename>",
|
| 54 |
+
"lstrip": false,
|
| 55 |
+
"normalized": false,
|
| 56 |
+
"rstrip": false,
|
| 57 |
+
"single_word": false,
|
| 58 |
+
"special": true
|
| 59 |
+
},
|
| 60 |
+
"7": {
|
| 61 |
+
"content": "<gh_stars>",
|
| 62 |
+
"lstrip": false,
|
| 63 |
+
"normalized": false,
|
| 64 |
+
"rstrip": false,
|
| 65 |
+
"single_word": false,
|
| 66 |
+
"special": true
|
| 67 |
+
},
|
| 68 |
+
"8": {
|
| 69 |
+
"content": "<issue_start>",
|
| 70 |
+
"lstrip": false,
|
| 71 |
+
"normalized": false,
|
| 72 |
+
"rstrip": false,
|
| 73 |
+
"single_word": false,
|
| 74 |
+
"special": true
|
| 75 |
+
},
|
| 76 |
+
"9": {
|
| 77 |
+
"content": "<issue_comment>",
|
| 78 |
+
"lstrip": false,
|
| 79 |
+
"normalized": false,
|
| 80 |
+
"rstrip": false,
|
| 81 |
+
"single_word": false,
|
| 82 |
+
"special": true
|
| 83 |
+
},
|
| 84 |
+
"10": {
|
| 85 |
+
"content": "<issue_closed>",
|
| 86 |
+
"lstrip": false,
|
| 87 |
+
"normalized": false,
|
| 88 |
+
"rstrip": false,
|
| 89 |
+
"single_word": false,
|
| 90 |
+
"special": true
|
| 91 |
+
},
|
| 92 |
+
"11": {
|
| 93 |
+
"content": "<jupyter_start>",
|
| 94 |
+
"lstrip": false,
|
| 95 |
+
"normalized": false,
|
| 96 |
+
"rstrip": false,
|
| 97 |
+
"single_word": false,
|
| 98 |
+
"special": true
|
| 99 |
+
},
|
| 100 |
+
"12": {
|
| 101 |
+
"content": "<jupyter_text>",
|
| 102 |
+
"lstrip": false,
|
| 103 |
+
"normalized": false,
|
| 104 |
+
"rstrip": false,
|
| 105 |
+
"single_word": false,
|
| 106 |
+
"special": true
|
| 107 |
+
},
|
| 108 |
+
"13": {
|
| 109 |
+
"content": "<jupyter_code>",
|
| 110 |
+
"lstrip": false,
|
| 111 |
+
"normalized": false,
|
| 112 |
+
"rstrip": false,
|
| 113 |
+
"single_word": false,
|
| 114 |
+
"special": true
|
| 115 |
+
},
|
| 116 |
+
"14": {
|
| 117 |
+
"content": "<jupyter_output>",
|
| 118 |
+
"lstrip": false,
|
| 119 |
+
"normalized": false,
|
| 120 |
+
"rstrip": false,
|
| 121 |
+
"single_word": false,
|
| 122 |
+
"special": true
|
| 123 |
+
},
|
| 124 |
+
"15": {
|
| 125 |
+
"content": "<jupyter_script>",
|
| 126 |
+
"lstrip": false,
|
| 127 |
+
"normalized": false,
|
| 128 |
+
"rstrip": false,
|
| 129 |
+
"single_word": false,
|
| 130 |
+
"special": true
|
| 131 |
+
},
|
| 132 |
+
"16": {
|
| 133 |
+
"content": "<empty_output>",
|
| 134 |
+
"lstrip": false,
|
| 135 |
+
"normalized": false,
|
| 136 |
+
"rstrip": false,
|
| 137 |
+
"single_word": false,
|
| 138 |
+
"special": true
|
| 139 |
+
}
|
| 140 |
+
},
|
| 141 |
+
"additional_special_tokens": [
|
| 142 |
+
"<|im_start|>",
|
| 143 |
+
"<|im_end|>"
|
| 144 |
+
],
|
| 145 |
+
"bos_token": "<|im_start|>",
|
| 146 |
+
"chat_template": "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful AI assistant named SmolLM, trained by Hugging Face<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
|
| 147 |
+
"clean_up_tokenization_spaces": false,
|
| 148 |
+
"eos_token": "<|im_end|>",
|
| 149 |
+
"extra_special_tokens": {},
|
| 150 |
+
"model_max_length": 8192,
|
| 151 |
+
"pad_token": "<|im_end|>",
|
| 152 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 153 |
+
"unk_token": "<|endoftext|>",
|
| 154 |
+
"vocab_size": 49152
|
| 155 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9997629770087698,
|
| 3 |
+
"total_flos": 141388749045760.0,
|
| 4 |
+
"train_loss": 0.6815844999777304,
|
| 5 |
+
"train_runtime": 3835.2311,
|
| 6 |
+
"train_samples": 270000,
|
| 7 |
+
"train_samples_per_second": 70.4,
|
| 8 |
+
"train_steps_per_second": 0.55
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2997 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9997629770087698,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2109,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.002370229912301493,
|
| 13 |
+
"grad_norm": 37.87957407825341,
|
| 14 |
+
"learning_rate": 7.109004739336492e-06,
|
| 15 |
+
"loss": 3.2574,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.004740459824602986,
|
| 20 |
+
"grad_norm": 18.177249571303804,
|
| 21 |
+
"learning_rate": 1.4218009478672984e-05,
|
| 22 |
+
"loss": 2.7918,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.00711068973690448,
|
| 27 |
+
"grad_norm": 5.001343795126334,
|
| 28 |
+
"learning_rate": 2.1327014218009475e-05,
|
| 29 |
+
"loss": 1.7693,
|
| 30 |
+
"step": 15
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.009480919649205973,
|
| 34 |
+
"grad_norm": 2.325367997984495,
|
| 35 |
+
"learning_rate": 2.8436018957345967e-05,
|
| 36 |
+
"loss": 1.2881,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.011851149561507466,
|
| 41 |
+
"grad_norm": 1.0235836184491791,
|
| 42 |
+
"learning_rate": 3.554502369668246e-05,
|
| 43 |
+
"loss": 1.0407,
|
| 44 |
+
"step": 25
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.01422137947380896,
|
| 48 |
+
"grad_norm": 0.6279074046022562,
|
| 49 |
+
"learning_rate": 4.265402843601895e-05,
|
| 50 |
+
"loss": 0.9531,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.016591609386110453,
|
| 55 |
+
"grad_norm": 0.6025558774407773,
|
| 56 |
+
"learning_rate": 4.9763033175355446e-05,
|
| 57 |
+
"loss": 0.8999,
|
| 58 |
+
"step": 35
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.018961839298411946,
|
| 62 |
+
"grad_norm": 0.4763893378885808,
|
| 63 |
+
"learning_rate": 5.6872037914691935e-05,
|
| 64 |
+
"loss": 0.8687,
|
| 65 |
+
"step": 40
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.02133206921071344,
|
| 69 |
+
"grad_norm": 0.37024070517127505,
|
| 70 |
+
"learning_rate": 6.398104265402843e-05,
|
| 71 |
+
"loss": 0.8436,
|
| 72 |
+
"step": 45
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.02370229912301493,
|
| 76 |
+
"grad_norm": 0.3630769483112416,
|
| 77 |
+
"learning_rate": 7.109004739336492e-05,
|
| 78 |
+
"loss": 0.8374,
|
| 79 |
+
"step": 50
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.026072529035316427,
|
| 83 |
+
"grad_norm": 0.20827819545344178,
|
| 84 |
+
"learning_rate": 7.819905213270142e-05,
|
| 85 |
+
"loss": 0.8075,
|
| 86 |
+
"step": 55
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.02844275894761792,
|
| 90 |
+
"grad_norm": 0.1601696569471494,
|
| 91 |
+
"learning_rate": 8.53080568720379e-05,
|
| 92 |
+
"loss": 0.7993,
|
| 93 |
+
"step": 60
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.030812988859919413,
|
| 97 |
+
"grad_norm": 0.1395108840386357,
|
| 98 |
+
"learning_rate": 9.24170616113744e-05,
|
| 99 |
+
"loss": 0.7844,
|
| 100 |
+
"step": 65
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.033183218772220906,
|
| 104 |
+
"grad_norm": 0.10908293522762928,
|
| 105 |
+
"learning_rate": 9.952606635071089e-05,
|
| 106 |
+
"loss": 0.7878,
|
| 107 |
+
"step": 70
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.0355534486845224,
|
| 111 |
+
"grad_norm": 0.09936997217620527,
|
| 112 |
+
"learning_rate": 0.00010663507109004738,
|
| 113 |
+
"loss": 0.7492,
|
| 114 |
+
"step": 75
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.03792367859682389,
|
| 118 |
+
"grad_norm": 0.09353265095940279,
|
| 119 |
+
"learning_rate": 0.00011374407582938387,
|
| 120 |
+
"loss": 0.7694,
|
| 121 |
+
"step": 80
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.040293908509125384,
|
| 125 |
+
"grad_norm": 0.10028988988229223,
|
| 126 |
+
"learning_rate": 0.00012085308056872037,
|
| 127 |
+
"loss": 0.7645,
|
| 128 |
+
"step": 85
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.04266413842142688,
|
| 132 |
+
"grad_norm": 0.08708652067409801,
|
| 133 |
+
"learning_rate": 0.00012796208530805686,
|
| 134 |
+
"loss": 0.7673,
|
| 135 |
+
"step": 90
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.04503436833372837,
|
| 139 |
+
"grad_norm": 0.07995303225473299,
|
| 140 |
+
"learning_rate": 0.00013507109004739335,
|
| 141 |
+
"loss": 0.7535,
|
| 142 |
+
"step": 95
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.04740459824602986,
|
| 146 |
+
"grad_norm": 0.07979293896147231,
|
| 147 |
+
"learning_rate": 0.00014218009478672984,
|
| 148 |
+
"loss": 0.7689,
|
| 149 |
+
"step": 100
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.049774828158331355,
|
| 153 |
+
"grad_norm": 0.08476243740502522,
|
| 154 |
+
"learning_rate": 0.00014928909952606633,
|
| 155 |
+
"loss": 0.7534,
|
| 156 |
+
"step": 105
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.052145058070632855,
|
| 160 |
+
"grad_norm": 0.07851254869100184,
|
| 161 |
+
"learning_rate": 0.00015639810426540285,
|
| 162 |
+
"loss": 0.7535,
|
| 163 |
+
"step": 110
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.05451528798293435,
|
| 167 |
+
"grad_norm": 0.08598983089548484,
|
| 168 |
+
"learning_rate": 0.0001635071090047393,
|
| 169 |
+
"loss": 0.7501,
|
| 170 |
+
"step": 115
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.05688551789523584,
|
| 174 |
+
"grad_norm": 0.08580450529497156,
|
| 175 |
+
"learning_rate": 0.0001706161137440758,
|
| 176 |
+
"loss": 0.7547,
|
| 177 |
+
"step": 120
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.05925574780753733,
|
| 181 |
+
"grad_norm": 0.08276852095286981,
|
| 182 |
+
"learning_rate": 0.00017772511848341231,
|
| 183 |
+
"loss": 0.7481,
|
| 184 |
+
"step": 125
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.061625977719838826,
|
| 188 |
+
"grad_norm": 0.0925763549038801,
|
| 189 |
+
"learning_rate": 0.0001848341232227488,
|
| 190 |
+
"loss": 0.7511,
|
| 191 |
+
"step": 130
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.06399620763214031,
|
| 195 |
+
"grad_norm": 0.07917042236946292,
|
| 196 |
+
"learning_rate": 0.00019194312796208527,
|
| 197 |
+
"loss": 0.752,
|
| 198 |
+
"step": 135
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.06636643754444181,
|
| 202 |
+
"grad_norm": 0.09521113296814705,
|
| 203 |
+
"learning_rate": 0.00019905213270142178,
|
| 204 |
+
"loss": 0.7525,
|
| 205 |
+
"step": 140
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.0687366674567433,
|
| 209 |
+
"grad_norm": 0.08361670279214221,
|
| 210 |
+
"learning_rate": 0.00020616113744075827,
|
| 211 |
+
"loss": 0.7499,
|
| 212 |
+
"step": 145
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.0711068973690448,
|
| 216 |
+
"grad_norm": 0.0808939093486955,
|
| 217 |
+
"learning_rate": 0.00021327014218009476,
|
| 218 |
+
"loss": 0.7564,
|
| 219 |
+
"step": 150
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.0734771272813463,
|
| 223 |
+
"grad_norm": 0.07840457880380175,
|
| 224 |
+
"learning_rate": 0.00022037914691943128,
|
| 225 |
+
"loss": 0.7456,
|
| 226 |
+
"step": 155
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.07584735719364778,
|
| 230 |
+
"grad_norm": 0.08285428720448265,
|
| 231 |
+
"learning_rate": 0.00022748815165876774,
|
| 232 |
+
"loss": 0.7573,
|
| 233 |
+
"step": 160
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.07821758710594928,
|
| 237 |
+
"grad_norm": 0.08834398679552397,
|
| 238 |
+
"learning_rate": 0.00023459715639810423,
|
| 239 |
+
"loss": 0.7528,
|
| 240 |
+
"step": 165
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.08058781701825077,
|
| 244 |
+
"grad_norm": 0.07707799033964063,
|
| 245 |
+
"learning_rate": 0.00024170616113744075,
|
| 246 |
+
"loss": 0.7555,
|
| 247 |
+
"step": 170
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.08295804693055227,
|
| 251 |
+
"grad_norm": 0.08232085673512189,
|
| 252 |
+
"learning_rate": 0.0002488151658767772,
|
| 253 |
+
"loss": 0.7428,
|
| 254 |
+
"step": 175
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.08532827684285375,
|
| 258 |
+
"grad_norm": 0.08887042007982729,
|
| 259 |
+
"learning_rate": 0.0002559241706161137,
|
| 260 |
+
"loss": 0.7624,
|
| 261 |
+
"step": 180
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.08769850675515525,
|
| 265 |
+
"grad_norm": 0.09470799851879333,
|
| 266 |
+
"learning_rate": 0.0002630331753554502,
|
| 267 |
+
"loss": 0.7528,
|
| 268 |
+
"step": 185
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.09006873666745674,
|
| 272 |
+
"grad_norm": 0.10657490417564752,
|
| 273 |
+
"learning_rate": 0.0002701421800947867,
|
| 274 |
+
"loss": 0.7557,
|
| 275 |
+
"step": 190
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.09243896657975824,
|
| 279 |
+
"grad_norm": 0.08182579598618632,
|
| 280 |
+
"learning_rate": 0.0002772511848341232,
|
| 281 |
+
"loss": 0.7364,
|
| 282 |
+
"step": 195
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.09480919649205972,
|
| 286 |
+
"grad_norm": 0.07272030097431592,
|
| 287 |
+
"learning_rate": 0.0002843601895734597,
|
| 288 |
+
"loss": 0.737,
|
| 289 |
+
"step": 200
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.09717942640436122,
|
| 293 |
+
"grad_norm": 0.0764754646270827,
|
| 294 |
+
"learning_rate": 0.0002914691943127962,
|
| 295 |
+
"loss": 0.7458,
|
| 296 |
+
"step": 205
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.09954965631666271,
|
| 300 |
+
"grad_norm": 0.08027094645738686,
|
| 301 |
+
"learning_rate": 0.00029857819905213266,
|
| 302 |
+
"loss": 0.7563,
|
| 303 |
+
"step": 210
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.10191988622896421,
|
| 307 |
+
"grad_norm": 0.08015350192584438,
|
| 308 |
+
"learning_rate": 0.0002999967123393108,
|
| 309 |
+
"loss": 0.7471,
|
| 310 |
+
"step": 215
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.10429011614126571,
|
| 314 |
+
"grad_norm": 0.07996649533232081,
|
| 315 |
+
"learning_rate": 0.0002999833564647566,
|
| 316 |
+
"loss": 0.7593,
|
| 317 |
+
"step": 220
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.1066603460535672,
|
| 321 |
+
"grad_norm": 0.07683862115921537,
|
| 322 |
+
"learning_rate": 0.00029995972781162177,
|
| 323 |
+
"loss": 0.7563,
|
| 324 |
+
"step": 225
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.1090305759658687,
|
| 328 |
+
"grad_norm": 0.08428065421678084,
|
| 329 |
+
"learning_rate": 0.0002999258279982986,
|
| 330 |
+
"loss": 0.7542,
|
| 331 |
+
"step": 230
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.11140080587817018,
|
| 335 |
+
"grad_norm": 0.09166260522899684,
|
| 336 |
+
"learning_rate": 0.00029988165934668,
|
| 337 |
+
"loss": 0.7422,
|
| 338 |
+
"step": 235
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.11377103579047168,
|
| 342 |
+
"grad_norm": 0.08294402646546879,
|
| 343 |
+
"learning_rate": 0.00029982722488200013,
|
| 344 |
+
"loss": 0.7464,
|
| 345 |
+
"step": 240
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.11614126570277317,
|
| 349 |
+
"grad_norm": 0.07426872103025435,
|
| 350 |
+
"learning_rate": 0.00029976252833262734,
|
| 351 |
+
"loss": 0.7568,
|
| 352 |
+
"step": 245
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.11851149561507467,
|
| 356 |
+
"grad_norm": 0.07034251693444554,
|
| 357 |
+
"learning_rate": 0.00029968757412980883,
|
| 358 |
+
"loss": 0.7515,
|
| 359 |
+
"step": 250
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.12088172552737615,
|
| 363 |
+
"grad_norm": 0.07550642047266844,
|
| 364 |
+
"learning_rate": 0.000299602367407367,
|
| 365 |
+
"loss": 0.7499,
|
| 366 |
+
"step": 255
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.12325195543967765,
|
| 370 |
+
"grad_norm": 0.07047433670873945,
|
| 371 |
+
"learning_rate": 0.00029950691400134824,
|
| 372 |
+
"loss": 0.7496,
|
| 373 |
+
"step": 260
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.12562218535197914,
|
| 377 |
+
"grad_norm": 0.0857398982188067,
|
| 378 |
+
"learning_rate": 0.0002994012204496225,
|
| 379 |
+
"loss": 0.7422,
|
| 380 |
+
"step": 265
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.12799241526428062,
|
| 384 |
+
"grad_norm": 0.08049042276394297,
|
| 385 |
+
"learning_rate": 0.00029928529399143606,
|
| 386 |
+
"loss": 0.7554,
|
| 387 |
+
"step": 270
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.13036264517658214,
|
| 391 |
+
"grad_norm": 0.08078197460014627,
|
| 392 |
+
"learning_rate": 0.0002991591425669156,
|
| 393 |
+
"loss": 0.7455,
|
| 394 |
+
"step": 275
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.13273287508888362,
|
| 398 |
+
"grad_norm": 0.07709419757715143,
|
| 399 |
+
"learning_rate": 0.0002990227748165241,
|
| 400 |
+
"loss": 0.7475,
|
| 401 |
+
"step": 280
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.1351031050011851,
|
| 405 |
+
"grad_norm": 0.07552451990431377,
|
| 406 |
+
"learning_rate": 0.00029887620008046934,
|
| 407 |
+
"loss": 0.7386,
|
| 408 |
+
"step": 285
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.1374733349134866,
|
| 412 |
+
"grad_norm": 0.07498431824773105,
|
| 413 |
+
"learning_rate": 0.00029871942839806396,
|
| 414 |
+
"loss": 0.7263,
|
| 415 |
+
"step": 290
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.1398435648257881,
|
| 419 |
+
"grad_norm": 44.58672438459057,
|
| 420 |
+
"learning_rate": 0.0002985524705070379,
|
| 421 |
+
"loss": 1.4646,
|
| 422 |
+
"step": 295
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.1422137947380896,
|
| 426 |
+
"grad_norm": 34.95300570734307,
|
| 427 |
+
"learning_rate": 0.0002983753378428031,
|
| 428 |
+
"loss": 3.0175,
|
| 429 |
+
"step": 300
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.14458402465039108,
|
| 433 |
+
"grad_norm": 1.898833557773242,
|
| 434 |
+
"learning_rate": 0.00029818804253766974,
|
| 435 |
+
"loss": 1.2782,
|
| 436 |
+
"step": 305
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.1469542545626926,
|
| 440 |
+
"grad_norm": 2.4268359613555375,
|
| 441 |
+
"learning_rate": 0.00029799059742001606,
|
| 442 |
+
"loss": 2.2389,
|
| 443 |
+
"step": 310
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.14932448447499408,
|
| 447 |
+
"grad_norm": 1.0640566578307415,
|
| 448 |
+
"learning_rate": 0.00029778301601340906,
|
| 449 |
+
"loss": 1.0058,
|
| 450 |
+
"step": 315
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.15169471438729557,
|
| 454 |
+
"grad_norm": 0.18597050864762904,
|
| 455 |
+
"learning_rate": 0.0002975653125356785,
|
| 456 |
+
"loss": 0.8664,
|
| 457 |
+
"step": 320
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.15406494429959705,
|
| 461 |
+
"grad_norm": 0.14018854346808488,
|
| 462 |
+
"learning_rate": 0.000297337501897943,
|
| 463 |
+
"loss": 0.8061,
|
| 464 |
+
"step": 325
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.15643517421189856,
|
| 468 |
+
"grad_norm": 0.1064736567889867,
|
| 469 |
+
"learning_rate": 0.0002970995997035888,
|
| 470 |
+
"loss": 0.7942,
|
| 471 |
+
"step": 330
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.15880540412420005,
|
| 475 |
+
"grad_norm": 0.09820103029046393,
|
| 476 |
+
"learning_rate": 0.0002968516222472011,
|
| 477 |
+
"loss": 0.7866,
|
| 478 |
+
"step": 335
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.16117563403650154,
|
| 482 |
+
"grad_norm": 0.10051042892221068,
|
| 483 |
+
"learning_rate": 0.0002965935865134476,
|
| 484 |
+
"loss": 0.7917,
|
| 485 |
+
"step": 340
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.16354586394880302,
|
| 489 |
+
"grad_norm": 0.07991174466381477,
|
| 490 |
+
"learning_rate": 0.0002963255101759159,
|
| 491 |
+
"loss": 0.7776,
|
| 492 |
+
"step": 345
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.16591609386110454,
|
| 496 |
+
"grad_norm": 0.07973784335269311,
|
| 497 |
+
"learning_rate": 0.0002960474115959023,
|
| 498 |
+
"loss": 0.7684,
|
| 499 |
+
"step": 350
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.16828632377340602,
|
| 503 |
+
"grad_norm": 3.2883719479596585,
|
| 504 |
+
"learning_rate": 0.00029575930982115445,
|
| 505 |
+
"loss": 0.7581,
|
| 506 |
+
"step": 355
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.1706565536857075,
|
| 510 |
+
"grad_norm": 0.31783313849101436,
|
| 511 |
+
"learning_rate": 0.0002954612245845669,
|
| 512 |
+
"loss": 1.0093,
|
| 513 |
+
"step": 360
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.17302678359800902,
|
| 517 |
+
"grad_norm": 0.14168083007400084,
|
| 518 |
+
"learning_rate": 0.0002951531763028292,
|
| 519 |
+
"loss": 0.8543,
|
| 520 |
+
"step": 365
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.1753970135103105,
|
| 524 |
+
"grad_norm": 0.11564628343820599,
|
| 525 |
+
"learning_rate": 0.0002948351860750277,
|
| 526 |
+
"loss": 0.8189,
|
| 527 |
+
"step": 370
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.177767243422612,
|
| 531 |
+
"grad_norm": 5.805978160293609,
|
| 532 |
+
"learning_rate": 0.00029450727568120037,
|
| 533 |
+
"loss": 0.8103,
|
| 534 |
+
"step": 375
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.18013747333491348,
|
| 538 |
+
"grad_norm": 0.08117200321663696,
|
| 539 |
+
"learning_rate": 0.0002941694675808451,
|
| 540 |
+
"loss": 0.7753,
|
| 541 |
+
"step": 380
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.182507703247215,
|
| 545 |
+
"grad_norm": 0.07607198738704968,
|
| 546 |
+
"learning_rate": 0.0002938217849113812,
|
| 547 |
+
"loss": 0.7603,
|
| 548 |
+
"step": 385
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.18487793315951648,
|
| 552 |
+
"grad_norm": 0.07113347917796854,
|
| 553 |
+
"learning_rate": 0.00029346425148656487,
|
| 554 |
+
"loss": 0.754,
|
| 555 |
+
"step": 390
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.18724816307181796,
|
| 559 |
+
"grad_norm": 0.06638185085999807,
|
| 560 |
+
"learning_rate": 0.0002930968917948581,
|
| 561 |
+
"loss": 0.7561,
|
| 562 |
+
"step": 395
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.18961839298411945,
|
| 566 |
+
"grad_norm": 0.0636288319721399,
|
| 567 |
+
"learning_rate": 0.000292719730997751,
|
| 568 |
+
"loss": 0.7431,
|
| 569 |
+
"step": 400
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.19198862289642096,
|
| 573 |
+
"grad_norm": 0.06951776019726015,
|
| 574 |
+
"learning_rate": 0.0002923327949280391,
|
| 575 |
+
"loss": 0.731,
|
| 576 |
+
"step": 405
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.19435885280872245,
|
| 580 |
+
"grad_norm": 0.0628405894324144,
|
| 581 |
+
"learning_rate": 0.0002919361100880534,
|
| 582 |
+
"loss": 0.735,
|
| 583 |
+
"step": 410
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.19672908272102393,
|
| 587 |
+
"grad_norm": 0.06251446911308753,
|
| 588 |
+
"learning_rate": 0.0002915297036478455,
|
| 589 |
+
"loss": 0.7277,
|
| 590 |
+
"step": 415
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.19909931263332542,
|
| 594 |
+
"grad_norm": 0.0606660234416891,
|
| 595 |
+
"learning_rate": 0.00029111360344332647,
|
| 596 |
+
"loss": 0.7324,
|
| 597 |
+
"step": 420
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.20146954254562693,
|
| 601 |
+
"grad_norm": 0.060370913004170065,
|
| 602 |
+
"learning_rate": 0.0002906878379743604,
|
| 603 |
+
"loss": 0.7234,
|
| 604 |
+
"step": 425
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.20383977245792842,
|
| 608 |
+
"grad_norm": 0.06265360740334025,
|
| 609 |
+
"learning_rate": 0.00029025243640281223,
|
| 610 |
+
"loss": 0.7357,
|
| 611 |
+
"step": 430
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.2062100023702299,
|
| 615 |
+
"grad_norm": 0.06358474472495813,
|
| 616 |
+
"learning_rate": 0.0002898074285505504,
|
| 617 |
+
"loss": 0.7188,
|
| 618 |
+
"step": 435
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.20858023228253142,
|
| 622 |
+
"grad_norm": 0.059826146661527135,
|
| 623 |
+
"learning_rate": 0.0002893528448974046,
|
| 624 |
+
"loss": 0.7183,
|
| 625 |
+
"step": 440
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.2109504621948329,
|
| 629 |
+
"grad_norm": 0.0601231927656965,
|
| 630 |
+
"learning_rate": 0.00028888871657907744,
|
| 631 |
+
"loss": 0.7189,
|
| 632 |
+
"step": 445
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.2133206921071344,
|
| 636 |
+
"grad_norm": 0.8526105639659928,
|
| 637 |
+
"learning_rate": 0.00028841507538501267,
|
| 638 |
+
"loss": 0.7202,
|
| 639 |
+
"step": 450
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.21569092201943588,
|
| 643 |
+
"grad_norm": 0.06361964303367311,
|
| 644 |
+
"learning_rate": 0.0002879319537562172,
|
| 645 |
+
"loss": 0.7191,
|
| 646 |
+
"step": 455
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.2180611519317374,
|
| 650 |
+
"grad_norm": 0.061558109294327805,
|
| 651 |
+
"learning_rate": 0.00028743938478303937,
|
| 652 |
+
"loss": 0.7202,
|
| 653 |
+
"step": 460
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.22043138184403888,
|
| 657 |
+
"grad_norm": 0.06228127012654291,
|
| 658 |
+
"learning_rate": 0.0002869374022029026,
|
| 659 |
+
"loss": 0.7269,
|
| 660 |
+
"step": 465
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.22280161175634036,
|
| 664 |
+
"grad_norm": 0.06239279055248459,
|
| 665 |
+
"learning_rate": 0.00028642604039799457,
|
| 666 |
+
"loss": 0.7211,
|
| 667 |
+
"step": 470
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.22517184166864185,
|
| 671 |
+
"grad_norm": 0.08371349247174907,
|
| 672 |
+
"learning_rate": 0.00028590533439291205,
|
| 673 |
+
"loss": 0.7132,
|
| 674 |
+
"step": 475
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.22754207158094336,
|
| 678 |
+
"grad_norm": 0.0630124916842619,
|
| 679 |
+
"learning_rate": 0.0002853753198522623,
|
| 680 |
+
"loss": 0.7216,
|
| 681 |
+
"step": 480
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.22991230149324485,
|
| 685 |
+
"grad_norm": 0.06231078679066273,
|
| 686 |
+
"learning_rate": 0.00028483603307822,
|
| 687 |
+
"loss": 0.7255,
|
| 688 |
+
"step": 485
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.23228253140554633,
|
| 692 |
+
"grad_norm": 0.06370990821920278,
|
| 693 |
+
"learning_rate": 0.00028428751100804115,
|
| 694 |
+
"loss": 0.7058,
|
| 695 |
+
"step": 490
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.23465276131784782,
|
| 699 |
+
"grad_norm": 0.05998852944044534,
|
| 700 |
+
"learning_rate": 0.0002837297912115329,
|
| 701 |
+
"loss": 0.7075,
|
| 702 |
+
"step": 495
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.23702299123014933,
|
| 706 |
+
"grad_norm": 0.05934170652874789,
|
| 707 |
+
"learning_rate": 0.00028316291188848025,
|
| 708 |
+
"loss": 0.6999,
|
| 709 |
+
"step": 500
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.23939322114245082,
|
| 713 |
+
"grad_norm": 0.06375843749865423,
|
| 714 |
+
"learning_rate": 0.0002825869118660298,
|
| 715 |
+
"loss": 0.713,
|
| 716 |
+
"step": 505
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.2417634510547523,
|
| 720 |
+
"grad_norm": 0.06159522236341869,
|
| 721 |
+
"learning_rate": 0.0002820018305960303,
|
| 722 |
+
"loss": 0.7088,
|
| 723 |
+
"step": 510
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.24413368096705382,
|
| 727 |
+
"grad_norm": 0.061758194634036964,
|
| 728 |
+
"learning_rate": 0.0002814077081523305,
|
| 729 |
+
"loss": 0.7042,
|
| 730 |
+
"step": 515
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.2465039108793553,
|
| 734 |
+
"grad_norm": 0.06170727020552632,
|
| 735 |
+
"learning_rate": 0.0002808045852280343,
|
| 736 |
+
"loss": 0.7103,
|
| 737 |
+
"step": 520
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.2488741407916568,
|
| 741 |
+
"grad_norm": 0.05840561215234994,
|
| 742 |
+
"learning_rate": 0.0002801925031327137,
|
| 743 |
+
"loss": 0.6987,
|
| 744 |
+
"step": 525
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.2512443707039583,
|
| 748 |
+
"grad_norm": 0.058175029378156634,
|
| 749 |
+
"learning_rate": 0.0002795715037895792,
|
| 750 |
+
"loss": 0.6909,
|
| 751 |
+
"step": 530
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.2536146006162598,
|
| 755 |
+
"grad_norm": 0.06101870086518658,
|
| 756 |
+
"learning_rate": 0.0002789416297326088,
|
| 757 |
+
"loss": 0.7184,
|
| 758 |
+
"step": 535
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.25598483052856125,
|
| 762 |
+
"grad_norm": 0.06279243933970817,
|
| 763 |
+
"learning_rate": 0.0002783029241036339,
|
| 764 |
+
"loss": 0.7123,
|
| 765 |
+
"step": 540
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.25835506044086276,
|
| 769 |
+
"grad_norm": 0.06346784165112655,
|
| 770 |
+
"learning_rate": 0.0002776554306493854,
|
| 771 |
+
"loss": 0.7196,
|
| 772 |
+
"step": 545
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.2607252903531643,
|
| 776 |
+
"grad_norm": 0.05748578268561143,
|
| 777 |
+
"learning_rate": 0.00027699919371849664,
|
| 778 |
+
"loss": 0.7084,
|
| 779 |
+
"step": 550
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.26309552026546573,
|
| 783 |
+
"grad_norm": 0.06563979843305176,
|
| 784 |
+
"learning_rate": 0.0002763342582584661,
|
| 785 |
+
"loss": 0.7048,
|
| 786 |
+
"step": 555
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.26546575017776725,
|
| 790 |
+
"grad_norm": 0.06177493929500722,
|
| 791 |
+
"learning_rate": 0.0002756606698125787,
|
| 792 |
+
"loss": 0.7058,
|
| 793 |
+
"step": 560
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.26783598009006876,
|
| 797 |
+
"grad_norm": 0.06067602378089436,
|
| 798 |
+
"learning_rate": 0.0002749784745167868,
|
| 799 |
+
"loss": 0.6934,
|
| 800 |
+
"step": 565
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.2702062100023702,
|
| 804 |
+
"grad_norm": 0.062834571657062,
|
| 805 |
+
"learning_rate": 0.0002742877190965495,
|
| 806 |
+
"loss": 0.6963,
|
| 807 |
+
"step": 570
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.27257643991467173,
|
| 811 |
+
"grad_norm": 0.059163687863750675,
|
| 812 |
+
"learning_rate": 0.0002735884508636331,
|
| 813 |
+
"loss": 0.7028,
|
| 814 |
+
"step": 575
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.2749466698269732,
|
| 818 |
+
"grad_norm": 0.05983910867748256,
|
| 819 |
+
"learning_rate": 0.00027288071771286996,
|
| 820 |
+
"loss": 0.6889,
|
| 821 |
+
"step": 580
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.2773168997392747,
|
| 825 |
+
"grad_norm": 0.05836025064363057,
|
| 826 |
+
"learning_rate": 0.0002721645681188784,
|
| 827 |
+
"loss": 0.6973,
|
| 828 |
+
"step": 585
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.2796871296515762,
|
| 832 |
+
"grad_norm": 0.06355277332945615,
|
| 833 |
+
"learning_rate": 0.00027144005113274237,
|
| 834 |
+
"loss": 0.6949,
|
| 835 |
+
"step": 590
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.2820573595638777,
|
| 839 |
+
"grad_norm": 0.060849753041127434,
|
| 840 |
+
"learning_rate": 0.0002707072163786518,
|
| 841 |
+
"loss": 0.69,
|
| 842 |
+
"step": 595
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.2844275894761792,
|
| 846 |
+
"grad_norm": 0.06170856816087761,
|
| 847 |
+
"learning_rate": 0.00026996611405050386,
|
| 848 |
+
"loss": 0.6976,
|
| 849 |
+
"step": 600
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.2867978193884807,
|
| 853 |
+
"grad_norm": 0.06086967149436421,
|
| 854 |
+
"learning_rate": 0.0002692167949084649,
|
| 855 |
+
"loss": 0.7081,
|
| 856 |
+
"step": 605
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.28916804930078216,
|
| 860 |
+
"grad_norm": 0.059663431709652645,
|
| 861 |
+
"learning_rate": 0.00026845931027549375,
|
| 862 |
+
"loss": 0.7034,
|
| 863 |
+
"step": 610
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.2915382792130837,
|
| 867 |
+
"grad_norm": 0.05768729651844958,
|
| 868 |
+
"learning_rate": 0.0002676937120338265,
|
| 869 |
+
"loss": 0.6988,
|
| 870 |
+
"step": 615
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.2939085091253852,
|
| 874 |
+
"grad_norm": 0.059449464935387684,
|
| 875 |
+
"learning_rate": 0.00026692005262142303,
|
| 876 |
+
"loss": 0.6911,
|
| 877 |
+
"step": 620
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.29627873903768664,
|
| 881 |
+
"grad_norm": 0.05908588685261821,
|
| 882 |
+
"learning_rate": 0.00026613838502837514,
|
| 883 |
+
"loss": 0.7012,
|
| 884 |
+
"step": 625
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.29864896894998816,
|
| 888 |
+
"grad_norm": 0.05823031421787097,
|
| 889 |
+
"learning_rate": 0.0002653487627932775,
|
| 890 |
+
"loss": 0.6815,
|
| 891 |
+
"step": 630
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.3010191988622896,
|
| 895 |
+
"grad_norm": 0.06346993001531062,
|
| 896 |
+
"learning_rate": 0.00026455123999956,
|
| 897 |
+
"loss": 0.6778,
|
| 898 |
+
"step": 635
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.30338942877459113,
|
| 902 |
+
"grad_norm": 0.05766157585577824,
|
| 903 |
+
"learning_rate": 0.00026374587127178434,
|
| 904 |
+
"loss": 0.6872,
|
| 905 |
+
"step": 640
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.30575965868689264,
|
| 909 |
+
"grad_norm": 0.060699498460503425,
|
| 910 |
+
"learning_rate": 0.00026293271177190184,
|
| 911 |
+
"loss": 0.698,
|
| 912 |
+
"step": 645
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.3081298885991941,
|
| 916 |
+
"grad_norm": 0.06470838061776032,
|
| 917 |
+
"learning_rate": 0.0002621118171954755,
|
| 918 |
+
"loss": 0.6863,
|
| 919 |
+
"step": 650
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.3105001185114956,
|
| 923 |
+
"grad_norm": 0.05809323133547399,
|
| 924 |
+
"learning_rate": 0.00026128324376786555,
|
| 925 |
+
"loss": 0.6853,
|
| 926 |
+
"step": 655
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.31287034842379713,
|
| 930 |
+
"grad_norm": 0.060605773782501364,
|
| 931 |
+
"learning_rate": 0.00026044704824037786,
|
| 932 |
+
"loss": 0.706,
|
| 933 |
+
"step": 660
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.3152405783360986,
|
| 937 |
+
"grad_norm": 0.05891479933624849,
|
| 938 |
+
"learning_rate": 0.0002596032878863774,
|
| 939 |
+
"loss": 0.6696,
|
| 940 |
+
"step": 665
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.3176108082484001,
|
| 944 |
+
"grad_norm": 0.0625427319419523,
|
| 945 |
+
"learning_rate": 0.0002587520204973651,
|
| 946 |
+
"loss": 0.6827,
|
| 947 |
+
"step": 670
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.3199810381607016,
|
| 951 |
+
"grad_norm": 0.06168949595593282,
|
| 952 |
+
"learning_rate": 0.00025789330437901974,
|
| 953 |
+
"loss": 0.6798,
|
| 954 |
+
"step": 675
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.32235126807300307,
|
| 958 |
+
"grad_norm": 0.05715106171580865,
|
| 959 |
+
"learning_rate": 0.00025702719834720433,
|
| 960 |
+
"loss": 0.6757,
|
| 961 |
+
"step": 680
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.3247214979853046,
|
| 965 |
+
"grad_norm": 0.05793783101869966,
|
| 966 |
+
"learning_rate": 0.00025615376172393774,
|
| 967 |
+
"loss": 0.6824,
|
| 968 |
+
"step": 685
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.32709172789760604,
|
| 972 |
+
"grad_norm": 0.05937949955225907,
|
| 973 |
+
"learning_rate": 0.00025527305433333127,
|
| 974 |
+
"loss": 0.6855,
|
| 975 |
+
"step": 690
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.32946195780990756,
|
| 979 |
+
"grad_norm": 0.06672193394439148,
|
| 980 |
+
"learning_rate": 0.00025438513649749173,
|
| 981 |
+
"loss": 0.6878,
|
| 982 |
+
"step": 695
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.33183218772220907,
|
| 986 |
+
"grad_norm": 0.06046441135071207,
|
| 987 |
+
"learning_rate": 0.00025349006903238906,
|
| 988 |
+
"loss": 0.6702,
|
| 989 |
+
"step": 700
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.33420241763451053,
|
| 993 |
+
"grad_norm": 0.06234393756385736,
|
| 994 |
+
"learning_rate": 0.0002525879132436916,
|
| 995 |
+
"loss": 0.6885,
|
| 996 |
+
"step": 705
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.33657264754681204,
|
| 1000 |
+
"grad_norm": 0.06442431249968464,
|
| 1001 |
+
"learning_rate": 0.0002516787309225664,
|
| 1002 |
+
"loss": 0.6824,
|
| 1003 |
+
"step": 710
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.33894287745911356,
|
| 1007 |
+
"grad_norm": 0.057449230900877835,
|
| 1008 |
+
"learning_rate": 0.0002507625843414477,
|
| 1009 |
+
"loss": 0.6825,
|
| 1010 |
+
"step": 715
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.341313107371415,
|
| 1014 |
+
"grad_norm": 0.06647075866637231,
|
| 1015 |
+
"learning_rate": 0.000249839536249771,
|
| 1016 |
+
"loss": 0.6715,
|
| 1017 |
+
"step": 720
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.34368333728371653,
|
| 1021 |
+
"grad_norm": 0.06516829971838099,
|
| 1022 |
+
"learning_rate": 0.0002489096498696758,
|
| 1023 |
+
"loss": 0.6618,
|
| 1024 |
+
"step": 725
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.34605356719601804,
|
| 1028 |
+
"grad_norm": 0.06010008501643476,
|
| 1029 |
+
"learning_rate": 0.0002479729888916749,
|
| 1030 |
+
"loss": 0.6791,
|
| 1031 |
+
"step": 730
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.3484237971083195,
|
| 1035 |
+
"grad_norm": 0.05742770698758986,
|
| 1036 |
+
"learning_rate": 0.0002470296174702921,
|
| 1037 |
+
"loss": 0.679,
|
| 1038 |
+
"step": 735
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.350794027020621,
|
| 1042 |
+
"grad_norm": 0.056906839817947766,
|
| 1043 |
+
"learning_rate": 0.0002460796002196684,
|
| 1044 |
+
"loss": 0.6667,
|
| 1045 |
+
"step": 740
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.35316425693292247,
|
| 1049 |
+
"grad_norm": 0.05683684204049849,
|
| 1050 |
+
"learning_rate": 0.00024512300220913617,
|
| 1051 |
+
"loss": 0.6771,
|
| 1052 |
+
"step": 745
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.355534486845224,
|
| 1056 |
+
"grad_norm": 0.05936901663811349,
|
| 1057 |
+
"learning_rate": 0.0002441598889587624,
|
| 1058 |
+
"loss": 0.6694,
|
| 1059 |
+
"step": 750
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.3579047167575255,
|
| 1063 |
+
"grad_norm": 0.056734740760335606,
|
| 1064 |
+
"learning_rate": 0.0002431903264348607,
|
| 1065 |
+
"loss": 0.6656,
|
| 1066 |
+
"step": 755
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.36027494666982696,
|
| 1070 |
+
"grad_norm": 0.05711271475645182,
|
| 1071 |
+
"learning_rate": 0.00024221438104547394,
|
| 1072 |
+
"loss": 0.6777,
|
| 1073 |
+
"step": 760
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.36264517658212847,
|
| 1077 |
+
"grad_norm": 0.06484350452592977,
|
| 1078 |
+
"learning_rate": 0.00024123211963582493,
|
| 1079 |
+
"loss": 0.6712,
|
| 1080 |
+
"step": 765
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.36501540649443,
|
| 1084 |
+
"grad_norm": 0.06427694084052618,
|
| 1085 |
+
"learning_rate": 0.00024024360948373834,
|
| 1086 |
+
"loss": 0.6655,
|
| 1087 |
+
"step": 770
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.36738563640673144,
|
| 1091 |
+
"grad_norm": 0.06329540165136863,
|
| 1092 |
+
"learning_rate": 0.0002392489182950328,
|
| 1093 |
+
"loss": 0.6649,
|
| 1094 |
+
"step": 775
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.36975586631903296,
|
| 1098 |
+
"grad_norm": 0.05829453649974294,
|
| 1099 |
+
"learning_rate": 0.0002382481141988833,
|
| 1100 |
+
"loss": 0.6825,
|
| 1101 |
+
"step": 780
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.3721260962313344,
|
| 1105 |
+
"grad_norm": 0.05785471731045562,
|
| 1106 |
+
"learning_rate": 0.00023724126574315484,
|
| 1107 |
+
"loss": 0.673,
|
| 1108 |
+
"step": 785
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.3744963261436359,
|
| 1112 |
+
"grad_norm": 0.0561890074327068,
|
| 1113 |
+
"learning_rate": 0.0002362284418897076,
|
| 1114 |
+
"loss": 0.6674,
|
| 1115 |
+
"step": 790
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.37686655605593744,
|
| 1119 |
+
"grad_norm": 0.05695140762918065,
|
| 1120 |
+
"learning_rate": 0.00023520971200967334,
|
| 1121 |
+
"loss": 0.6699,
|
| 1122 |
+
"step": 795
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.3792367859682389,
|
| 1126 |
+
"grad_norm": 0.056382588072258655,
|
| 1127 |
+
"learning_rate": 0.00023418514587870414,
|
| 1128 |
+
"loss": 0.6669,
|
| 1129 |
+
"step": 800
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.3816070158805404,
|
| 1133 |
+
"grad_norm": 0.05699605706668468,
|
| 1134 |
+
"learning_rate": 0.00023315481367219306,
|
| 1135 |
+
"loss": 0.6646,
|
| 1136 |
+
"step": 805
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.3839772457928419,
|
| 1140 |
+
"grad_norm": 0.05971904780219733,
|
| 1141 |
+
"learning_rate": 0.00023211878596046795,
|
| 1142 |
+
"loss": 0.6694,
|
| 1143 |
+
"step": 810
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.3863474757051434,
|
| 1147 |
+
"grad_norm": 0.05872202455985711,
|
| 1148 |
+
"learning_rate": 0.00023107713370395752,
|
| 1149 |
+
"loss": 0.6564,
|
| 1150 |
+
"step": 815
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.3887177056174449,
|
| 1154 |
+
"grad_norm": 0.056267010708901014,
|
| 1155 |
+
"learning_rate": 0.00023002992824833134,
|
| 1156 |
+
"loss": 0.6619,
|
| 1157 |
+
"step": 820
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.3910879355297464,
|
| 1161 |
+
"grad_norm": 0.05858819926921814,
|
| 1162 |
+
"learning_rate": 0.00022897724131961314,
|
| 1163 |
+
"loss": 0.6555,
|
| 1164 |
+
"step": 825
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.39345816544204787,
|
| 1168 |
+
"grad_norm": 0.058908341888103316,
|
| 1169 |
+
"learning_rate": 0.00022791914501926785,
|
| 1170 |
+
"loss": 0.6587,
|
| 1171 |
+
"step": 830
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.3958283953543494,
|
| 1175 |
+
"grad_norm": 0.05581026158289953,
|
| 1176 |
+
"learning_rate": 0.00022685571181926354,
|
| 1177 |
+
"loss": 0.6725,
|
| 1178 |
+
"step": 835
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.39819862526665084,
|
| 1182 |
+
"grad_norm": 0.05851912322437669,
|
| 1183 |
+
"learning_rate": 0.00022578701455710734,
|
| 1184 |
+
"loss": 0.6627,
|
| 1185 |
+
"step": 840
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.40056885517895235,
|
| 1189 |
+
"grad_norm": 0.05512523024478408,
|
| 1190 |
+
"learning_rate": 0.00022471312643085667,
|
| 1191 |
+
"loss": 0.6597,
|
| 1192 |
+
"step": 845
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.40293908509125387,
|
| 1196 |
+
"grad_norm": 0.05488965997035471,
|
| 1197 |
+
"learning_rate": 0.00022363412099410585,
|
| 1198 |
+
"loss": 0.6546,
|
| 1199 |
+
"step": 850
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.4053093150035553,
|
| 1203 |
+
"grad_norm": 0.05590348015503589,
|
| 1204 |
+
"learning_rate": 0.0002225500721509479,
|
| 1205 |
+
"loss": 0.6676,
|
| 1206 |
+
"step": 855
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.40767954491585684,
|
| 1210 |
+
"grad_norm": 0.0557504280408415,
|
| 1211 |
+
"learning_rate": 0.0002214610541509129,
|
| 1212 |
+
"loss": 0.6559,
|
| 1213 |
+
"step": 860
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.41004977482815835,
|
| 1217 |
+
"grad_norm": 0.060424783129262576,
|
| 1218 |
+
"learning_rate": 0.00022036714158388235,
|
| 1219 |
+
"loss": 0.6712,
|
| 1220 |
+
"step": 865
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.4124200047404598,
|
| 1224 |
+
"grad_norm": 0.05481897236135071,
|
| 1225 |
+
"learning_rate": 0.00021926840937498034,
|
| 1226 |
+
"loss": 0.6506,
|
| 1227 |
+
"step": 870
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.4147902346527613,
|
| 1231 |
+
"grad_norm": 0.05656177938481429,
|
| 1232 |
+
"learning_rate": 0.00021816493277944157,
|
| 1233 |
+
"loss": 0.6746,
|
| 1234 |
+
"step": 875
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.41716046456506284,
|
| 1238 |
+
"grad_norm": 0.05625948208637603,
|
| 1239 |
+
"learning_rate": 0.00021705678737745713,
|
| 1240 |
+
"loss": 0.6378,
|
| 1241 |
+
"step": 880
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.4195306944773643,
|
| 1245 |
+
"grad_norm": 0.05903464854176921,
|
| 1246 |
+
"learning_rate": 0.0002159440490689976,
|
| 1247 |
+
"loss": 0.6606,
|
| 1248 |
+
"step": 885
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.4219009243896658,
|
| 1252 |
+
"grad_norm": 0.05683800007835736,
|
| 1253 |
+
"learning_rate": 0.0002148267940686146,
|
| 1254 |
+
"loss": 0.6621,
|
| 1255 |
+
"step": 890
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.42427115430196727,
|
| 1259 |
+
"grad_norm": 0.05931947846319525,
|
| 1260 |
+
"learning_rate": 0.0002137050989002206,
|
| 1261 |
+
"loss": 0.6539,
|
| 1262 |
+
"step": 895
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.4266413842142688,
|
| 1266 |
+
"grad_norm": 0.05335168668707516,
|
| 1267 |
+
"learning_rate": 0.0002125790403918475,
|
| 1268 |
+
"loss": 0.656,
|
| 1269 |
+
"step": 900
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.4290116141265703,
|
| 1273 |
+
"grad_norm": 0.0544878383415537,
|
| 1274 |
+
"learning_rate": 0.0002114486956703846,
|
| 1275 |
+
"loss": 0.6438,
|
| 1276 |
+
"step": 905
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.43138184403887175,
|
| 1280 |
+
"grad_norm": 0.06212920385348046,
|
| 1281 |
+
"learning_rate": 0.00021031414215629576,
|
| 1282 |
+
"loss": 0.6519,
|
| 1283 |
+
"step": 910
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.43375207395117327,
|
| 1287 |
+
"grad_norm": 0.053801804211920747,
|
| 1288 |
+
"learning_rate": 0.000209175457558317,
|
| 1289 |
+
"loss": 0.6534,
|
| 1290 |
+
"step": 915
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.4361223038634748,
|
| 1294 |
+
"grad_norm": 0.055676827319867436,
|
| 1295 |
+
"learning_rate": 0.00020803271986813364,
|
| 1296 |
+
"loss": 0.6434,
|
| 1297 |
+
"step": 920
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.43849253377577624,
|
| 1301 |
+
"grad_norm": 0.06221867524670911,
|
| 1302 |
+
"learning_rate": 0.00020688600735503863,
|
| 1303 |
+
"loss": 0.6642,
|
| 1304 |
+
"step": 925
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.44086276368807775,
|
| 1308 |
+
"grad_norm": 0.05719745436620134,
|
| 1309 |
+
"learning_rate": 0.0002057353985605717,
|
| 1310 |
+
"loss": 0.6654,
|
| 1311 |
+
"step": 930
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.4432329936003792,
|
| 1315 |
+
"grad_norm": 0.05636204685752076,
|
| 1316 |
+
"learning_rate": 0.00020458097229313963,
|
| 1317 |
+
"loss": 0.6469,
|
| 1318 |
+
"step": 935
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.4456032235126807,
|
| 1322 |
+
"grad_norm": 0.05949021529775284,
|
| 1323 |
+
"learning_rate": 0.00020342280762261874,
|
| 1324 |
+
"loss": 0.6514,
|
| 1325 |
+
"step": 940
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.44797345342498224,
|
| 1329 |
+
"grad_norm": 0.05819024554018189,
|
| 1330 |
+
"learning_rate": 0.0002022609838749389,
|
| 1331 |
+
"loss": 0.6601,
|
| 1332 |
+
"step": 945
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.4503436833372837,
|
| 1336 |
+
"grad_norm": 0.054298063737541105,
|
| 1337 |
+
"learning_rate": 0.00020109558062665037,
|
| 1338 |
+
"loss": 0.6561,
|
| 1339 |
+
"step": 950
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.4527139132495852,
|
| 1343 |
+
"grad_norm": 0.05325573283656968,
|
| 1344 |
+
"learning_rate": 0.00019992667769947348,
|
| 1345 |
+
"loss": 0.6556,
|
| 1346 |
+
"step": 955
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.4550841431618867,
|
| 1350 |
+
"grad_norm": 0.055557610793003705,
|
| 1351 |
+
"learning_rate": 0.00019875435515483126,
|
| 1352 |
+
"loss": 0.6598,
|
| 1353 |
+
"step": 960
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.4574543730741882,
|
| 1357 |
+
"grad_norm": 0.05410922961948251,
|
| 1358 |
+
"learning_rate": 0.0001975786932883659,
|
| 1359 |
+
"loss": 0.6494,
|
| 1360 |
+
"step": 965
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.4598246029864897,
|
| 1364 |
+
"grad_norm": 0.05513130388145586,
|
| 1365 |
+
"learning_rate": 0.00019639977262443908,
|
| 1366 |
+
"loss": 0.6458,
|
| 1367 |
+
"step": 970
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.4621948328987912,
|
| 1371 |
+
"grad_norm": 0.05722322905223257,
|
| 1372 |
+
"learning_rate": 0.00019521767391061663,
|
| 1373 |
+
"loss": 0.6708,
|
| 1374 |
+
"step": 975
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.46456506281109267,
|
| 1378 |
+
"grad_norm": 0.06292159556421471,
|
| 1379 |
+
"learning_rate": 0.00019403247811213776,
|
| 1380 |
+
"loss": 0.6482,
|
| 1381 |
+
"step": 980
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.4669352927233942,
|
| 1385 |
+
"grad_norm": 0.061013794703927336,
|
| 1386 |
+
"learning_rate": 0.00019284426640636982,
|
| 1387 |
+
"loss": 0.6407,
|
| 1388 |
+
"step": 985
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.46930552263569564,
|
| 1392 |
+
"grad_norm": 0.05521810364010896,
|
| 1393 |
+
"learning_rate": 0.000191653120177248,
|
| 1394 |
+
"loss": 0.6433,
|
| 1395 |
+
"step": 990
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.47167575254799715,
|
| 1399 |
+
"grad_norm": 0.05464230308163921,
|
| 1400 |
+
"learning_rate": 0.00019045912100970112,
|
| 1401 |
+
"loss": 0.6408,
|
| 1402 |
+
"step": 995
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.47404598246029866,
|
| 1406 |
+
"grad_norm": 0.05534936230890589,
|
| 1407 |
+
"learning_rate": 0.00018926235068406378,
|
| 1408 |
+
"loss": 0.6406,
|
| 1409 |
+
"step": 1000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.4764162123726001,
|
| 1413 |
+
"grad_norm": 0.05541859380358046,
|
| 1414 |
+
"learning_rate": 0.00018806289117047494,
|
| 1415 |
+
"loss": 0.6453,
|
| 1416 |
+
"step": 1005
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.47878644228490164,
|
| 1420 |
+
"grad_norm": 0.057019399656362324,
|
| 1421 |
+
"learning_rate": 0.00018686082462326346,
|
| 1422 |
+
"loss": 0.6424,
|
| 1423 |
+
"step": 1010
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.48115667219720315,
|
| 1427 |
+
"grad_norm": 0.05678834933291726,
|
| 1428 |
+
"learning_rate": 0.00018565623337532137,
|
| 1429 |
+
"loss": 0.6372,
|
| 1430 |
+
"step": 1015
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.4835269021095046,
|
| 1434 |
+
"grad_norm": 0.0626356752316691,
|
| 1435 |
+
"learning_rate": 0.00018444919993246427,
|
| 1436 |
+
"loss": 0.6536,
|
| 1437 |
+
"step": 1020
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.4858971320218061,
|
| 1441 |
+
"grad_norm": 0.05533208427640317,
|
| 1442 |
+
"learning_rate": 0.00018323980696778074,
|
| 1443 |
+
"loss": 0.6392,
|
| 1444 |
+
"step": 1025
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.48826736193410764,
|
| 1448 |
+
"grad_norm": 0.1583885749203545,
|
| 1449 |
+
"learning_rate": 0.0001820281373159694,
|
| 1450 |
+
"loss": 0.6377,
|
| 1451 |
+
"step": 1030
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.4906375918464091,
|
| 1455 |
+
"grad_norm": 0.05696942950681657,
|
| 1456 |
+
"learning_rate": 0.00018081427396766565,
|
| 1457 |
+
"loss": 0.632,
|
| 1458 |
+
"step": 1035
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.4930078217587106,
|
| 1462 |
+
"grad_norm": 0.06885427735472549,
|
| 1463 |
+
"learning_rate": 0.0001795983000637572,
|
| 1464 |
+
"loss": 0.6513,
|
| 1465 |
+
"step": 1040
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.49537805167101207,
|
| 1469 |
+
"grad_norm": 0.055666281836575904,
|
| 1470 |
+
"learning_rate": 0.00017838029888968973,
|
| 1471 |
+
"loss": 0.64,
|
| 1472 |
+
"step": 1045
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.4977482815833136,
|
| 1476 |
+
"grad_norm": 0.05590039370646873,
|
| 1477 |
+
"learning_rate": 0.00017716035386976215,
|
| 1478 |
+
"loss": 0.6363,
|
| 1479 |
+
"step": 1050
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.500118511495615,
|
| 1483 |
+
"grad_norm": 0.05747509683831957,
|
| 1484 |
+
"learning_rate": 0.00017593854856141287,
|
| 1485 |
+
"loss": 0.622,
|
| 1486 |
+
"step": 1055
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.5024887414079166,
|
| 1490 |
+
"grad_norm": 0.05599406050839353,
|
| 1491 |
+
"learning_rate": 0.00017471496664949666,
|
| 1492 |
+
"loss": 0.6352,
|
| 1493 |
+
"step": 1060
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.5048589713202181,
|
| 1497 |
+
"grad_norm": 0.056394714786088526,
|
| 1498 |
+
"learning_rate": 0.00017348969194055285,
|
| 1499 |
+
"loss": 0.628,
|
| 1500 |
+
"step": 1065
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.5072292012325196,
|
| 1504 |
+
"grad_norm": 0.05379996817708946,
|
| 1505 |
+
"learning_rate": 0.00017226280835706503,
|
| 1506 |
+
"loss": 0.6253,
|
| 1507 |
+
"step": 1070
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.5095994311448211,
|
| 1511 |
+
"grad_norm": 0.052644081064009085,
|
| 1512 |
+
"learning_rate": 0.00017103439993171326,
|
| 1513 |
+
"loss": 0.635,
|
| 1514 |
+
"step": 1075
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.5119696610571225,
|
| 1518 |
+
"grad_norm": 0.06183603107070903,
|
| 1519 |
+
"learning_rate": 0.00016980455080161817,
|
| 1520 |
+
"loss": 0.6247,
|
| 1521 |
+
"step": 1080
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.514339890969424,
|
| 1525 |
+
"grad_norm": 0.055776407065979555,
|
| 1526 |
+
"learning_rate": 0.00016857334520257837,
|
| 1527 |
+
"loss": 0.6356,
|
| 1528 |
+
"step": 1085
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.5167101208817255,
|
| 1532 |
+
"grad_norm": 0.057457278964487705,
|
| 1533 |
+
"learning_rate": 0.00016734086746330066,
|
| 1534 |
+
"loss": 0.6253,
|
| 1535 |
+
"step": 1090
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.519080350794027,
|
| 1539 |
+
"grad_norm": 0.054238874177121164,
|
| 1540 |
+
"learning_rate": 0.00016610720199962444,
|
| 1541 |
+
"loss": 0.6394,
|
| 1542 |
+
"step": 1095
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.5214505807063285,
|
| 1546 |
+
"grad_norm": 0.05556319159032643,
|
| 1547 |
+
"learning_rate": 0.00016487243330873957,
|
| 1548 |
+
"loss": 0.647,
|
| 1549 |
+
"step": 1100
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.5238208106186301,
|
| 1553 |
+
"grad_norm": 0.05533410752493418,
|
| 1554 |
+
"learning_rate": 0.00016363664596339905,
|
| 1555 |
+
"loss": 0.6301,
|
| 1556 |
+
"step": 1105
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.5261910405309315,
|
| 1560 |
+
"grad_norm": 0.057938893164981804,
|
| 1561 |
+
"learning_rate": 0.00016239992460612636,
|
| 1562 |
+
"loss": 0.6229,
|
| 1563 |
+
"step": 1110
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.528561270443233,
|
| 1567 |
+
"grad_norm": 0.05173600564638707,
|
| 1568 |
+
"learning_rate": 0.00016116235394341807,
|
| 1569 |
+
"loss": 0.6277,
|
| 1570 |
+
"step": 1115
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.5309315003555345,
|
| 1574 |
+
"grad_norm": 0.055079698685841325,
|
| 1575 |
+
"learning_rate": 0.00015992401873994206,
|
| 1576 |
+
"loss": 0.6169,
|
| 1577 |
+
"step": 1120
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.533301730267836,
|
| 1581 |
+
"grad_norm": 0.05549655687980898,
|
| 1582 |
+
"learning_rate": 0.00015868500381273167,
|
| 1583 |
+
"loss": 0.6389,
|
| 1584 |
+
"step": 1125
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.5356719601801375,
|
| 1588 |
+
"grad_norm": 0.059015018470196176,
|
| 1589 |
+
"learning_rate": 0.0001574453940253765,
|
| 1590 |
+
"loss": 0.6406,
|
| 1591 |
+
"step": 1130
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.5380421900924389,
|
| 1595 |
+
"grad_norm": 0.05635084046958829,
|
| 1596 |
+
"learning_rate": 0.0001562052742822098,
|
| 1597 |
+
"loss": 0.6245,
|
| 1598 |
+
"step": 1135
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.5404124200047404,
|
| 1602 |
+
"grad_norm": 0.056720749371614865,
|
| 1603 |
+
"learning_rate": 0.00015496472952249298,
|
| 1604 |
+
"loss": 0.6322,
|
| 1605 |
+
"step": 1140
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.542782649917042,
|
| 1609 |
+
"grad_norm": 0.055644623371899404,
|
| 1610 |
+
"learning_rate": 0.00015372384471459816,
|
| 1611 |
+
"loss": 0.6338,
|
| 1612 |
+
"step": 1145
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.5451528798293435,
|
| 1616 |
+
"grad_norm": 0.056962336114346236,
|
| 1617 |
+
"learning_rate": 0.0001524827048501883,
|
| 1618 |
+
"loss": 0.6363,
|
| 1619 |
+
"step": 1150
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.547523109741645,
|
| 1623 |
+
"grad_norm": 3.4539173382130626,
|
| 1624 |
+
"learning_rate": 0.00015124139493839575,
|
| 1625 |
+
"loss": 0.6514,
|
| 1626 |
+
"step": 1155
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.5498933396539464,
|
| 1630 |
+
"grad_norm": 0.0592035753126842,
|
| 1631 |
+
"learning_rate": 0.00015,
|
| 1632 |
+
"loss": 0.6354,
|
| 1633 |
+
"step": 1160
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.5522635695662479,
|
| 1637 |
+
"grad_norm": 0.05848608337431566,
|
| 1638 |
+
"learning_rate": 0.0001487586050616043,
|
| 1639 |
+
"loss": 0.6291,
|
| 1640 |
+
"step": 1165
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.5546337994785494,
|
| 1644 |
+
"grad_norm": 0.056837303715594346,
|
| 1645 |
+
"learning_rate": 0.00014751729514981169,
|
| 1646 |
+
"loss": 0.6186,
|
| 1647 |
+
"step": 1170
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.5570040293908509,
|
| 1651 |
+
"grad_norm": 0.05430228685183931,
|
| 1652 |
+
"learning_rate": 0.0001462761552854018,
|
| 1653 |
+
"loss": 0.6353,
|
| 1654 |
+
"step": 1175
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.5593742593031524,
|
| 1658 |
+
"grad_norm": 0.05561193103292618,
|
| 1659 |
+
"learning_rate": 0.00014503527047750702,
|
| 1660 |
+
"loss": 0.621,
|
| 1661 |
+
"step": 1180
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.561744489215454,
|
| 1665 |
+
"grad_norm": 0.054843289485073124,
|
| 1666 |
+
"learning_rate": 0.00014379472571779024,
|
| 1667 |
+
"loss": 0.6142,
|
| 1668 |
+
"step": 1185
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.5641147191277553,
|
| 1672 |
+
"grad_norm": 0.05315905080273884,
|
| 1673 |
+
"learning_rate": 0.00014255460597462346,
|
| 1674 |
+
"loss": 0.6161,
|
| 1675 |
+
"step": 1190
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.5664849490400569,
|
| 1679 |
+
"grad_norm": 0.05493581701875387,
|
| 1680 |
+
"learning_rate": 0.00014131499618726833,
|
| 1681 |
+
"loss": 0.6285,
|
| 1682 |
+
"step": 1195
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.5688551789523584,
|
| 1686 |
+
"grad_norm": 0.05700162567210758,
|
| 1687 |
+
"learning_rate": 0.00014007598126005797,
|
| 1688 |
+
"loss": 0.6257,
|
| 1689 |
+
"step": 1200
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.5712254088646599,
|
| 1693 |
+
"grad_norm": 0.052067584151166996,
|
| 1694 |
+
"learning_rate": 0.0001388376460565819,
|
| 1695 |
+
"loss": 0.6141,
|
| 1696 |
+
"step": 1205
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.5735956387769614,
|
| 1700 |
+
"grad_norm": 0.057269236575854914,
|
| 1701 |
+
"learning_rate": 0.00013760007539387364,
|
| 1702 |
+
"loss": 0.626,
|
| 1703 |
+
"step": 1210
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.5759658686892628,
|
| 1707 |
+
"grad_norm": 0.05865861625327605,
|
| 1708 |
+
"learning_rate": 0.00013636335403660092,
|
| 1709 |
+
"loss": 0.6275,
|
| 1710 |
+
"step": 1215
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.5783360986015643,
|
| 1714 |
+
"grad_norm": 0.055532429030743166,
|
| 1715 |
+
"learning_rate": 0.00013512756669126043,
|
| 1716 |
+
"loss": 0.6205,
|
| 1717 |
+
"step": 1220
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.5807063285138658,
|
| 1721 |
+
"grad_norm": 0.06249705139064605,
|
| 1722 |
+
"learning_rate": 0.00013389279800037553,
|
| 1723 |
+
"loss": 0.6343,
|
| 1724 |
+
"step": 1225
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.5830765584261673,
|
| 1728 |
+
"grad_norm": 0.053614563419465064,
|
| 1729 |
+
"learning_rate": 0.0001326591325366993,
|
| 1730 |
+
"loss": 0.6113,
|
| 1731 |
+
"step": 1230
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.5854467883384689,
|
| 1735 |
+
"grad_norm": 0.05424894433907924,
|
| 1736 |
+
"learning_rate": 0.00013142665479742163,
|
| 1737 |
+
"loss": 0.623,
|
| 1738 |
+
"step": 1235
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.5878170182507704,
|
| 1742 |
+
"grad_norm": 0.0542014909020612,
|
| 1743 |
+
"learning_rate": 0.00013019544919838177,
|
| 1744 |
+
"loss": 0.6186,
|
| 1745 |
+
"step": 1240
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.5901872481630718,
|
| 1749 |
+
"grad_norm": 0.053382059431365064,
|
| 1750 |
+
"learning_rate": 0.0001289656000682867,
|
| 1751 |
+
"loss": 0.607,
|
| 1752 |
+
"step": 1245
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.5925574780753733,
|
| 1756 |
+
"grad_norm": 0.05601301326761515,
|
| 1757 |
+
"learning_rate": 0.00012773719164293497,
|
| 1758 |
+
"loss": 0.6185,
|
| 1759 |
+
"step": 1250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.5949277079876748,
|
| 1763 |
+
"grad_norm": 0.05319350407228568,
|
| 1764 |
+
"learning_rate": 0.00012651030805944713,
|
| 1765 |
+
"loss": 0.6093,
|
| 1766 |
+
"step": 1255
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.5972979378999763,
|
| 1770 |
+
"grad_norm": 0.05115446334747276,
|
| 1771 |
+
"learning_rate": 0.0001252850333505033,
|
| 1772 |
+
"loss": 0.615,
|
| 1773 |
+
"step": 1260
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.5996681678122778,
|
| 1777 |
+
"grad_norm": 0.05192578571658311,
|
| 1778 |
+
"learning_rate": 0.00012406145143858713,
|
| 1779 |
+
"loss": 0.6136,
|
| 1780 |
+
"step": 1265
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.6020383977245792,
|
| 1784 |
+
"grad_norm": 0.054594901119558795,
|
| 1785 |
+
"learning_rate": 0.00012283964613023788,
|
| 1786 |
+
"loss": 0.6053,
|
| 1787 |
+
"step": 1270
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.6044086276368807,
|
| 1791 |
+
"grad_norm": 0.055180387588156125,
|
| 1792 |
+
"learning_rate": 0.00012161970111031023,
|
| 1793 |
+
"loss": 0.6235,
|
| 1794 |
+
"step": 1275
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.6067788575491823,
|
| 1798 |
+
"grad_norm": 0.05162312778472747,
|
| 1799 |
+
"learning_rate": 0.00012040169993624276,
|
| 1800 |
+
"loss": 0.6224,
|
| 1801 |
+
"step": 1280
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.6091490874614838,
|
| 1805 |
+
"grad_norm": 0.05862655622704016,
|
| 1806 |
+
"learning_rate": 0.00011918572603233436,
|
| 1807 |
+
"loss": 0.6208,
|
| 1808 |
+
"step": 1285
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.6115193173737853,
|
| 1812 |
+
"grad_norm": 0.05577369996114947,
|
| 1813 |
+
"learning_rate": 0.00011797186268403057,
|
| 1814 |
+
"loss": 0.6176,
|
| 1815 |
+
"step": 1290
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.6138895472860868,
|
| 1819 |
+
"grad_norm": 0.05380234343377544,
|
| 1820 |
+
"learning_rate": 0.00011676019303221925,
|
| 1821 |
+
"loss": 0.6133,
|
| 1822 |
+
"step": 1295
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.6162597771983882,
|
| 1826 |
+
"grad_norm": 0.058884284873815654,
|
| 1827 |
+
"learning_rate": 0.0001155508000675357,
|
| 1828 |
+
"loss": 0.6053,
|
| 1829 |
+
"step": 1300
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.6186300071106897,
|
| 1833 |
+
"grad_norm": 0.053535760742643955,
|
| 1834 |
+
"learning_rate": 0.00011434376662467867,
|
| 1835 |
+
"loss": 0.6142,
|
| 1836 |
+
"step": 1305
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.6210002370229912,
|
| 1840 |
+
"grad_norm": 0.05339335441771012,
|
| 1841 |
+
"learning_rate": 0.00011313917537673648,
|
| 1842 |
+
"loss": 0.6069,
|
| 1843 |
+
"step": 1310
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.6233704669352927,
|
| 1847 |
+
"grad_norm": 0.05379638473884841,
|
| 1848 |
+
"learning_rate": 0.00011193710882952506,
|
| 1849 |
+
"loss": 0.5951,
|
| 1850 |
+
"step": 1315
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.6257406968475943,
|
| 1854 |
+
"grad_norm": 0.05461329438051225,
|
| 1855 |
+
"learning_rate": 0.00011073764931593622,
|
| 1856 |
+
"loss": 0.6236,
|
| 1857 |
+
"step": 1320
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.6281109267598957,
|
| 1861 |
+
"grad_norm": 0.056457300525490715,
|
| 1862 |
+
"learning_rate": 0.00010954087899029885,
|
| 1863 |
+
"loss": 0.6056,
|
| 1864 |
+
"step": 1325
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.6304811566721972,
|
| 1868 |
+
"grad_norm": 0.05737998956682243,
|
| 1869 |
+
"learning_rate": 0.00010834687982275199,
|
| 1870 |
+
"loss": 0.6023,
|
| 1871 |
+
"step": 1330
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.6328513865844987,
|
| 1875 |
+
"grad_norm": 0.05393272686617499,
|
| 1876 |
+
"learning_rate": 0.00010715573359363018,
|
| 1877 |
+
"loss": 0.607,
|
| 1878 |
+
"step": 1335
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.6352216164968002,
|
| 1882 |
+
"grad_norm": 0.058430747172485496,
|
| 1883 |
+
"learning_rate": 0.00010596752188786223,
|
| 1884 |
+
"loss": 0.6145,
|
| 1885 |
+
"step": 1340
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.6375918464091017,
|
| 1889 |
+
"grad_norm": 0.05380330186102191,
|
| 1890 |
+
"learning_rate": 0.00010478232608938339,
|
| 1891 |
+
"loss": 0.6038,
|
| 1892 |
+
"step": 1345
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.6399620763214032,
|
| 1896 |
+
"grad_norm": 0.057322372930785785,
|
| 1897 |
+
"learning_rate": 0.0001036002273755609,
|
| 1898 |
+
"loss": 0.6081,
|
| 1899 |
+
"step": 1350
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.6423323062337046,
|
| 1903 |
+
"grad_norm": 0.059264371354861206,
|
| 1904 |
+
"learning_rate": 0.0001024213067116341,
|
| 1905 |
+
"loss": 0.6007,
|
| 1906 |
+
"step": 1355
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.6447025361460061,
|
| 1910 |
+
"grad_norm": 0.05436742717666459,
|
| 1911 |
+
"learning_rate": 0.00010124564484516874,
|
| 1912 |
+
"loss": 0.614,
|
| 1913 |
+
"step": 1360
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.6470727660583077,
|
| 1917 |
+
"grad_norm": 0.055704527406812185,
|
| 1918 |
+
"learning_rate": 0.00010007332230052652,
|
| 1919 |
+
"loss": 0.611,
|
| 1920 |
+
"step": 1365
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.6494429959706092,
|
| 1924 |
+
"grad_norm": 0.05458191321816212,
|
| 1925 |
+
"learning_rate": 9.890441937334963e-05,
|
| 1926 |
+
"loss": 0.6018,
|
| 1927 |
+
"step": 1370
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.6518132258829107,
|
| 1931 |
+
"grad_norm": 0.05555570625108052,
|
| 1932 |
+
"learning_rate": 9.773901612506113e-05,
|
| 1933 |
+
"loss": 0.5986,
|
| 1934 |
+
"step": 1375
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.6541834557952121,
|
| 1938 |
+
"grad_norm": 0.05694735645213942,
|
| 1939 |
+
"learning_rate": 9.657719237738122e-05,
|
| 1940 |
+
"loss": 0.6161,
|
| 1941 |
+
"step": 1380
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.6565536857075136,
|
| 1945 |
+
"grad_norm": 0.05412760351903458,
|
| 1946 |
+
"learning_rate": 9.541902770686034e-05,
|
| 1947 |
+
"loss": 0.5933,
|
| 1948 |
+
"step": 1385
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.6589239156198151,
|
| 1952 |
+
"grad_norm": 0.05413679325940457,
|
| 1953 |
+
"learning_rate": 9.426460143942832e-05,
|
| 1954 |
+
"loss": 0.6037,
|
| 1955 |
+
"step": 1390
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.6612941455321166,
|
| 1959 |
+
"grad_norm": 0.05174840431523283,
|
| 1960 |
+
"learning_rate": 9.311399264496135e-05,
|
| 1961 |
+
"loss": 0.5947,
|
| 1962 |
+
"step": 1395
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.6636643754444181,
|
| 1966 |
+
"grad_norm": 0.0532643825199599,
|
| 1967 |
+
"learning_rate": 9.196728013186636e-05,
|
| 1968 |
+
"loss": 0.5998,
|
| 1969 |
+
"step": 1400
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.6660346053567197,
|
| 1973 |
+
"grad_norm": 0.055311945512636886,
|
| 1974 |
+
"learning_rate": 9.082454244168301e-05,
|
| 1975 |
+
"loss": 0.5961,
|
| 1976 |
+
"step": 1405
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.6684048352690211,
|
| 1980 |
+
"grad_norm": 0.05305987298812406,
|
| 1981 |
+
"learning_rate": 8.968585784370424e-05,
|
| 1982 |
+
"loss": 0.6057,
|
| 1983 |
+
"step": 1410
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.6707750651813226,
|
| 1987 |
+
"grad_norm": 0.051852358998889965,
|
| 1988 |
+
"learning_rate": 8.855130432961541e-05,
|
| 1989 |
+
"loss": 0.5847,
|
| 1990 |
+
"step": 1415
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.6731452950936241,
|
| 1994 |
+
"grad_norm": 0.05695676096470251,
|
| 1995 |
+
"learning_rate": 8.74209596081525e-05,
|
| 1996 |
+
"loss": 0.6027,
|
| 1997 |
+
"step": 1420
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.6755155250059256,
|
| 2001 |
+
"grad_norm": 0.052927987042953824,
|
| 2002 |
+
"learning_rate": 8.629490109977937e-05,
|
| 2003 |
+
"loss": 0.6149,
|
| 2004 |
+
"step": 1425
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.6778857549182271,
|
| 2008 |
+
"grad_norm": 0.05418803856885555,
|
| 2009 |
+
"learning_rate": 8.517320593138534e-05,
|
| 2010 |
+
"loss": 0.6056,
|
| 2011 |
+
"step": 1430
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.6802559848305285,
|
| 2015 |
+
"grad_norm": 0.05515959054343932,
|
| 2016 |
+
"learning_rate": 8.405595093100234e-05,
|
| 2017 |
+
"loss": 0.6057,
|
| 2018 |
+
"step": 1435
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.68262621474283,
|
| 2022 |
+
"grad_norm": 0.054880537220212015,
|
| 2023 |
+
"learning_rate": 8.294321262254287e-05,
|
| 2024 |
+
"loss": 0.5905,
|
| 2025 |
+
"step": 1440
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.6849964446551315,
|
| 2029 |
+
"grad_norm": 0.054943521074427744,
|
| 2030 |
+
"learning_rate": 8.183506722055836e-05,
|
| 2031 |
+
"loss": 0.5945,
|
| 2032 |
+
"step": 1445
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.6873666745674331,
|
| 2036 |
+
"grad_norm": 0.05472729142594704,
|
| 2037 |
+
"learning_rate": 8.073159062501963e-05,
|
| 2038 |
+
"loss": 0.5926,
|
| 2039 |
+
"step": 1450
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.6897369044797346,
|
| 2043 |
+
"grad_norm": 0.058850442027131224,
|
| 2044 |
+
"learning_rate": 7.963285841611758e-05,
|
| 2045 |
+
"loss": 0.6008,
|
| 2046 |
+
"step": 1455
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.6921071343920361,
|
| 2050 |
+
"grad_norm": 0.05707369096484443,
|
| 2051 |
+
"learning_rate": 7.853894584908711e-05,
|
| 2052 |
+
"loss": 0.6039,
|
| 2053 |
+
"step": 1460
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.6944773643043375,
|
| 2057 |
+
"grad_norm": 0.0523587623400406,
|
| 2058 |
+
"learning_rate": 7.744992784905206e-05,
|
| 2059 |
+
"loss": 0.6007,
|
| 2060 |
+
"step": 1465
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.696847594216639,
|
| 2064 |
+
"grad_norm": 0.053321883103446634,
|
| 2065 |
+
"learning_rate": 7.636587900589415e-05,
|
| 2066 |
+
"loss": 0.5856,
|
| 2067 |
+
"step": 1470
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.6992178241289405,
|
| 2071 |
+
"grad_norm": 0.05442178077056066,
|
| 2072 |
+
"learning_rate": 7.52868735691433e-05,
|
| 2073 |
+
"loss": 0.6024,
|
| 2074 |
+
"step": 1475
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.701588054041242,
|
| 2078 |
+
"grad_norm": 0.05378553098177081,
|
| 2079 |
+
"learning_rate": 7.421298544289267e-05,
|
| 2080 |
+
"loss": 0.6041,
|
| 2081 |
+
"step": 1480
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.7039582839535435,
|
| 2085 |
+
"grad_norm": 0.05107025701826348,
|
| 2086 |
+
"learning_rate": 7.314428818073645e-05,
|
| 2087 |
+
"loss": 0.5978,
|
| 2088 |
+
"step": 1485
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.7063285138658449,
|
| 2092 |
+
"grad_norm": 0.056885586353233,
|
| 2093 |
+
"learning_rate": 7.208085498073218e-05,
|
| 2094 |
+
"loss": 0.5982,
|
| 2095 |
+
"step": 1490
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.7086987437781465,
|
| 2099 |
+
"grad_norm": 0.05367334237421415,
|
| 2100 |
+
"learning_rate": 7.102275868038689e-05,
|
| 2101 |
+
"loss": 0.5938,
|
| 2102 |
+
"step": 1495
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.711068973690448,
|
| 2106 |
+
"grad_norm": 0.049920747527540395,
|
| 2107 |
+
"learning_rate": 6.997007175166862e-05,
|
| 2108 |
+
"loss": 0.5801,
|
| 2109 |
+
"step": 1500
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.7134392036027495,
|
| 2113 |
+
"grad_norm": 0.053158788899790074,
|
| 2114 |
+
"learning_rate": 6.892286629604244e-05,
|
| 2115 |
+
"loss": 0.5867,
|
| 2116 |
+
"step": 1505
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.715809433515051,
|
| 2120 |
+
"grad_norm": 0.05623068287636579,
|
| 2121 |
+
"learning_rate": 6.788121403953207e-05,
|
| 2122 |
+
"loss": 0.5958,
|
| 2123 |
+
"step": 1510
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.7181796634273524,
|
| 2127 |
+
"grad_norm": 0.05310094430866351,
|
| 2128 |
+
"learning_rate": 6.684518632780688e-05,
|
| 2129 |
+
"loss": 0.5936,
|
| 2130 |
+
"step": 1515
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.7205498933396539,
|
| 2134 |
+
"grad_norm": 0.051694922971736335,
|
| 2135 |
+
"learning_rate": 6.581485412129587e-05,
|
| 2136 |
+
"loss": 0.5893,
|
| 2137 |
+
"step": 1520
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.7229201232519554,
|
| 2141 |
+
"grad_norm": 0.05405965263345783,
|
| 2142 |
+
"learning_rate": 6.479028799032664e-05,
|
| 2143 |
+
"loss": 0.603,
|
| 2144 |
+
"step": 1525
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.7252903531642569,
|
| 2148 |
+
"grad_norm": 0.05541132610068022,
|
| 2149 |
+
"learning_rate": 6.37715581102924e-05,
|
| 2150 |
+
"loss": 0.5887,
|
| 2151 |
+
"step": 1530
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.7276605830765585,
|
| 2155 |
+
"grad_norm": 0.05286163558669179,
|
| 2156 |
+
"learning_rate": 6.275873425684514e-05,
|
| 2157 |
+
"loss": 0.5881,
|
| 2158 |
+
"step": 1535
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.73003081298886,
|
| 2162 |
+
"grad_norm": 0.052024799630321195,
|
| 2163 |
+
"learning_rate": 6.175188580111667e-05,
|
| 2164 |
+
"loss": 0.6045,
|
| 2165 |
+
"step": 1540
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.7324010429011614,
|
| 2169 |
+
"grad_norm": 0.055373729325846666,
|
| 2170 |
+
"learning_rate": 6.075108170496718e-05,
|
| 2171 |
+
"loss": 0.5832,
|
| 2172 |
+
"step": 1545
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.7347712728134629,
|
| 2176 |
+
"grad_norm": 0.05516378115033556,
|
| 2177 |
+
"learning_rate": 5.9756390516261576e-05,
|
| 2178 |
+
"loss": 0.5934,
|
| 2179 |
+
"step": 1550
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.7371415027257644,
|
| 2183 |
+
"grad_norm": 0.05458895386019013,
|
| 2184 |
+
"learning_rate": 5.8767880364175076e-05,
|
| 2185 |
+
"loss": 0.6055,
|
| 2186 |
+
"step": 1555
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.7395117326380659,
|
| 2190 |
+
"grad_norm": 0.054173922532646177,
|
| 2191 |
+
"learning_rate": 5.778561895452602e-05,
|
| 2192 |
+
"loss": 0.5912,
|
| 2193 |
+
"step": 1560
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.7418819625503674,
|
| 2197 |
+
"grad_norm": 0.0535683841018571,
|
| 2198 |
+
"learning_rate": 5.680967356513931e-05,
|
| 2199 |
+
"loss": 0.5846,
|
| 2200 |
+
"step": 1565
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.7442521924626688,
|
| 2204 |
+
"grad_norm": 0.05450496516039284,
|
| 2205 |
+
"learning_rate": 5.5840111041237614e-05,
|
| 2206 |
+
"loss": 0.5813,
|
| 2207 |
+
"step": 1570
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.7466224223749703,
|
| 2211 |
+
"grad_norm": 0.054578000158860905,
|
| 2212 |
+
"learning_rate": 5.48769977908638e-05,
|
| 2213 |
+
"loss": 0.593,
|
| 2214 |
+
"step": 1575
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.7489926522872719,
|
| 2218 |
+
"grad_norm": 0.0515712714221276,
|
| 2219 |
+
"learning_rate": 5.392039978033157e-05,
|
| 2220 |
+
"loss": 0.5927,
|
| 2221 |
+
"step": 1580
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.7513628821995734,
|
| 2225 |
+
"grad_norm": 0.05321518991804319,
|
| 2226 |
+
"learning_rate": 5.29703825297079e-05,
|
| 2227 |
+
"loss": 0.5831,
|
| 2228 |
+
"step": 1585
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.7537331121118749,
|
| 2232 |
+
"grad_norm": 0.05298981424430984,
|
| 2233 |
+
"learning_rate": 5.202701110832513e-05,
|
| 2234 |
+
"loss": 0.5892,
|
| 2235 |
+
"step": 1590
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.7561033420241764,
|
| 2239 |
+
"grad_norm": 0.052907289573775475,
|
| 2240 |
+
"learning_rate": 5.1090350130324196e-05,
|
| 2241 |
+
"loss": 0.5943,
|
| 2242 |
+
"step": 1595
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.7584735719364778,
|
| 2246 |
+
"grad_norm": 0.05190800487566414,
|
| 2247 |
+
"learning_rate": 5.016046375022892e-05,
|
| 2248 |
+
"loss": 0.5802,
|
| 2249 |
+
"step": 1600
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.7608438018487793,
|
| 2253 |
+
"grad_norm": 0.0538931538504174,
|
| 2254 |
+
"learning_rate": 4.923741565855231e-05,
|
| 2255 |
+
"loss": 0.5818,
|
| 2256 |
+
"step": 1605
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.7632140317610808,
|
| 2260 |
+
"grad_norm": 0.052494006253503485,
|
| 2261 |
+
"learning_rate": 4.832126907743356e-05,
|
| 2262 |
+
"loss": 0.5861,
|
| 2263 |
+
"step": 1610
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.7655842616733823,
|
| 2267 |
+
"grad_norm": 0.05335836205095365,
|
| 2268 |
+
"learning_rate": 4.7412086756308456e-05,
|
| 2269 |
+
"loss": 0.5834,
|
| 2270 |
+
"step": 1615
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.7679544915856839,
|
| 2274 |
+
"grad_norm": 0.05509114267028979,
|
| 2275 |
+
"learning_rate": 4.650993096761088e-05,
|
| 2276 |
+
"loss": 0.5869,
|
| 2277 |
+
"step": 1620
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.7703247214979853,
|
| 2281 |
+
"grad_norm": 0.052187260630405116,
|
| 2282 |
+
"learning_rate": 4.5614863502508274e-05,
|
| 2283 |
+
"loss": 0.6069,
|
| 2284 |
+
"step": 1625
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.7726949514102868,
|
| 2288 |
+
"grad_norm": 0.05431858623792003,
|
| 2289 |
+
"learning_rate": 4.4726945666668675e-05,
|
| 2290 |
+
"loss": 0.5946,
|
| 2291 |
+
"step": 1630
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.7750651813225883,
|
| 2295 |
+
"grad_norm": 0.052723752996540384,
|
| 2296 |
+
"learning_rate": 4.384623827606225e-05,
|
| 2297 |
+
"loss": 0.5789,
|
| 2298 |
+
"step": 1635
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.7774354112348898,
|
| 2302 |
+
"grad_norm": 0.05256027199703368,
|
| 2303 |
+
"learning_rate": 4.297280165279562e-05,
|
| 2304 |
+
"loss": 0.5716,
|
| 2305 |
+
"step": 1640
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.7798056411471913,
|
| 2309 |
+
"grad_norm": 0.054172042789112265,
|
| 2310 |
+
"learning_rate": 4.210669562098021e-05,
|
| 2311 |
+
"loss": 0.5873,
|
| 2312 |
+
"step": 1645
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.7821758710594928,
|
| 2316 |
+
"grad_norm": 0.05964267895580253,
|
| 2317 |
+
"learning_rate": 4.124797950263491e-05,
|
| 2318 |
+
"loss": 0.5883,
|
| 2319 |
+
"step": 1650
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.7845461009717942,
|
| 2323 |
+
"grad_norm": 0.052843311666916505,
|
| 2324 |
+
"learning_rate": 4.039671211362262e-05,
|
| 2325 |
+
"loss": 0.5912,
|
| 2326 |
+
"step": 1655
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.7869163308840957,
|
| 2330 |
+
"grad_norm": 0.05799110456913064,
|
| 2331 |
+
"learning_rate": 3.9552951759622134e-05,
|
| 2332 |
+
"loss": 0.5836,
|
| 2333 |
+
"step": 1660
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.7892865607963973,
|
| 2337 |
+
"grad_norm": 0.053733688999035426,
|
| 2338 |
+
"learning_rate": 3.8716756232134436e-05,
|
| 2339 |
+
"loss": 0.5873,
|
| 2340 |
+
"step": 1665
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.7916567907086988,
|
| 2344 |
+
"grad_norm": 0.05503743591767426,
|
| 2345 |
+
"learning_rate": 3.788818280452443e-05,
|
| 2346 |
+
"loss": 0.5806,
|
| 2347 |
+
"step": 1670
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.7940270206210003,
|
| 2351 |
+
"grad_norm": 0.05314889782998826,
|
| 2352 |
+
"learning_rate": 3.706728822809813e-05,
|
| 2353 |
+
"loss": 0.5859,
|
| 2354 |
+
"step": 1675
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.7963972505333017,
|
| 2358 |
+
"grad_norm": 0.05263806721046257,
|
| 2359 |
+
"learning_rate": 3.625412872821565e-05,
|
| 2360 |
+
"loss": 0.5701,
|
| 2361 |
+
"step": 1680
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.7987674804456032,
|
| 2365 |
+
"grad_norm": 0.053056629300361355,
|
| 2366 |
+
"learning_rate": 3.544876000043997e-05,
|
| 2367 |
+
"loss": 0.5689,
|
| 2368 |
+
"step": 1685
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.8011377103579047,
|
| 2372 |
+
"grad_norm": 0.05166101520125829,
|
| 2373 |
+
"learning_rate": 3.465123720672255e-05,
|
| 2374 |
+
"loss": 0.5624,
|
| 2375 |
+
"step": 1690
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.8035079402702062,
|
| 2379 |
+
"grad_norm": 0.05341566401567655,
|
| 2380 |
+
"learning_rate": 3.3861614971624825e-05,
|
| 2381 |
+
"loss": 0.597,
|
| 2382 |
+
"step": 1695
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.8058781701825077,
|
| 2386 |
+
"grad_norm": 0.05184611062114667,
|
| 2387 |
+
"learning_rate": 3.3079947378577e-05,
|
| 2388 |
+
"loss": 0.589,
|
| 2389 |
+
"step": 1700
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.8082484000948092,
|
| 2393 |
+
"grad_norm": 0.05652651621528694,
|
| 2394 |
+
"learning_rate": 3.230628796617345e-05,
|
| 2395 |
+
"loss": 0.5656,
|
| 2396 |
+
"step": 1705
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.8106186300071107,
|
| 2400 |
+
"grad_norm": 0.05185062445049837,
|
| 2401 |
+
"learning_rate": 3.154068972450626e-05,
|
| 2402 |
+
"loss": 0.5803,
|
| 2403 |
+
"step": 1710
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.8129888599194122,
|
| 2407 |
+
"grad_norm": 0.05236594549663459,
|
| 2408 |
+
"learning_rate": 3.07832050915351e-05,
|
| 2409 |
+
"loss": 0.5687,
|
| 2410 |
+
"step": 1715
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.8153590898317137,
|
| 2414 |
+
"grad_norm": 0.052981202994182464,
|
| 2415 |
+
"learning_rate": 3.0033885949496157e-05,
|
| 2416 |
+
"loss": 0.5845,
|
| 2417 |
+
"step": 1720
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.8177293197440152,
|
| 2421 |
+
"grad_norm": 0.05119910241673355,
|
| 2422 |
+
"learning_rate": 2.9292783621348176e-05,
|
| 2423 |
+
"loss": 0.5881,
|
| 2424 |
+
"step": 1725
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.8200995496563167,
|
| 2428 |
+
"grad_norm": 0.05198844865138733,
|
| 2429 |
+
"learning_rate": 2.8559948867257643e-05,
|
| 2430 |
+
"loss": 0.5733,
|
| 2431 |
+
"step": 1730
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.8224697795686181,
|
| 2435 |
+
"grad_norm": 0.0664972668514182,
|
| 2436 |
+
"learning_rate": 2.783543188112156e-05,
|
| 2437 |
+
"loss": 0.5748,
|
| 2438 |
+
"step": 1735
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.8248400094809196,
|
| 2442 |
+
"grad_norm": 0.05135611458630472,
|
| 2443 |
+
"learning_rate": 2.711928228712999e-05,
|
| 2444 |
+
"loss": 0.5649,
|
| 2445 |
+
"step": 1740
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 0.8272102393932211,
|
| 2449 |
+
"grad_norm": 0.05397408303013281,
|
| 2450 |
+
"learning_rate": 2.6411549136366874e-05,
|
| 2451 |
+
"loss": 0.5889,
|
| 2452 |
+
"step": 1745
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.8295804693055226,
|
| 2456 |
+
"grad_norm": 0.052454038077671426,
|
| 2457 |
+
"learning_rate": 2.5712280903450495e-05,
|
| 2458 |
+
"loss": 0.6,
|
| 2459 |
+
"step": 1750
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.8319506992178242,
|
| 2463 |
+
"grad_norm": 0.05123417132112601,
|
| 2464 |
+
"learning_rate": 2.502152548321324e-05,
|
| 2465 |
+
"loss": 0.5707,
|
| 2466 |
+
"step": 1755
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 0.8343209291301257,
|
| 2470 |
+
"grad_norm": 0.05251617509320771,
|
| 2471 |
+
"learning_rate": 2.4339330187421264e-05,
|
| 2472 |
+
"loss": 0.5769,
|
| 2473 |
+
"step": 1760
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.8366911590424271,
|
| 2477 |
+
"grad_norm": 0.06318258557079645,
|
| 2478 |
+
"learning_rate": 2.3665741741533918e-05,
|
| 2479 |
+
"loss": 0.5727,
|
| 2480 |
+
"step": 1765
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 0.8390613889547286,
|
| 2484 |
+
"grad_norm": 0.05419644476733012,
|
| 2485 |
+
"learning_rate": 2.3000806281503324e-05,
|
| 2486 |
+
"loss": 0.5639,
|
| 2487 |
+
"step": 1770
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 0.8414316188670301,
|
| 2491 |
+
"grad_norm": 0.0529717737403971,
|
| 2492 |
+
"learning_rate": 2.2344569350614548e-05,
|
| 2493 |
+
"loss": 0.5747,
|
| 2494 |
+
"step": 1775
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 0.8438018487793316,
|
| 2498 |
+
"grad_norm": 0.053183308485240915,
|
| 2499 |
+
"learning_rate": 2.169707589636606e-05,
|
| 2500 |
+
"loss": 0.5756,
|
| 2501 |
+
"step": 1780
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.8461720786916331,
|
| 2505 |
+
"grad_norm": 0.048981198328497996,
|
| 2506 |
+
"learning_rate": 2.1058370267391245e-05,
|
| 2507 |
+
"loss": 0.5594,
|
| 2508 |
+
"step": 1785
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.8485423086039345,
|
| 2512 |
+
"grad_norm": 0.047709819338697244,
|
| 2513 |
+
"learning_rate": 2.0428496210420753e-05,
|
| 2514 |
+
"loss": 0.5814,
|
| 2515 |
+
"step": 1790
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.850912538516236,
|
| 2519 |
+
"grad_norm": 0.051806025941571895,
|
| 2520 |
+
"learning_rate": 1.980749686728631e-05,
|
| 2521 |
+
"loss": 0.5853,
|
| 2522 |
+
"step": 1795
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 0.8532827684285376,
|
| 2526 |
+
"grad_norm": 0.05238919786303432,
|
| 2527 |
+
"learning_rate": 1.919541477196569e-05,
|
| 2528 |
+
"loss": 0.5793,
|
| 2529 |
+
"step": 1800
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.8556529983408391,
|
| 2533 |
+
"grad_norm": 0.050957261348074434,
|
| 2534 |
+
"learning_rate": 1.859229184766951e-05,
|
| 2535 |
+
"loss": 0.5724,
|
| 2536 |
+
"step": 1805
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 0.8580232282531406,
|
| 2540 |
+
"grad_norm": 0.052439653700969896,
|
| 2541 |
+
"learning_rate": 1.7998169403969652e-05,
|
| 2542 |
+
"loss": 0.5794,
|
| 2543 |
+
"step": 1810
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.860393458165442,
|
| 2547 |
+
"grad_norm": 0.05158899579880472,
|
| 2548 |
+
"learning_rate": 1.741308813397017e-05,
|
| 2549 |
+
"loss": 0.587,
|
| 2550 |
+
"step": 1815
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.8627636880777435,
|
| 2554 |
+
"grad_norm": 0.05347972640720942,
|
| 2555 |
+
"learning_rate": 1.683708811151972e-05,
|
| 2556 |
+
"loss": 0.5737,
|
| 2557 |
+
"step": 1820
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.865133917990045,
|
| 2561 |
+
"grad_norm": 0.05251163235177713,
|
| 2562 |
+
"learning_rate": 1.6270208788467054e-05,
|
| 2563 |
+
"loss": 0.5851,
|
| 2564 |
+
"step": 1825
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.8675041479023465,
|
| 2568 |
+
"grad_norm": 0.05130891776047808,
|
| 2569 |
+
"learning_rate": 1.5712488991958773e-05,
|
| 2570 |
+
"loss": 0.5653,
|
| 2571 |
+
"step": 1830
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 0.869874377814648,
|
| 2575 |
+
"grad_norm": 0.051624628347411626,
|
| 2576 |
+
"learning_rate": 1.5163966921779974e-05,
|
| 2577 |
+
"loss": 0.5631,
|
| 2578 |
+
"step": 1835
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 0.8722446077269496,
|
| 2582 |
+
"grad_norm": 0.05192898094836235,
|
| 2583 |
+
"learning_rate": 1.4624680147737694e-05,
|
| 2584 |
+
"loss": 0.5701,
|
| 2585 |
+
"step": 1840
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.874614837639251,
|
| 2589 |
+
"grad_norm": 0.05273604182501612,
|
| 2590 |
+
"learning_rate": 1.4094665607087919e-05,
|
| 2591 |
+
"loss": 0.5739,
|
| 2592 |
+
"step": 1845
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.8769850675515525,
|
| 2596 |
+
"grad_norm": 0.05064520739634672,
|
| 2597 |
+
"learning_rate": 1.357395960200538e-05,
|
| 2598 |
+
"loss": 0.5692,
|
| 2599 |
+
"step": 1850
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.879355297463854,
|
| 2603 |
+
"grad_norm": 0.050839315021426905,
|
| 2604 |
+
"learning_rate": 1.3062597797097346e-05,
|
| 2605 |
+
"loss": 0.5704,
|
| 2606 |
+
"step": 1855
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 0.8817255273761555,
|
| 2610 |
+
"grad_norm": 0.051292554412449225,
|
| 2611 |
+
"learning_rate": 1.2560615216960607e-05,
|
| 2612 |
+
"loss": 0.5695,
|
| 2613 |
+
"step": 1860
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 0.884095757288457,
|
| 2617 |
+
"grad_norm": 0.05149899783415366,
|
| 2618 |
+
"learning_rate": 1.2068046243782826e-05,
|
| 2619 |
+
"loss": 0.5708,
|
| 2620 |
+
"step": 1865
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.8864659872007584,
|
| 2624 |
+
"grad_norm": 0.05155770982273923,
|
| 2625 |
+
"learning_rate": 1.1584924614987323e-05,
|
| 2626 |
+
"loss": 0.5682,
|
| 2627 |
+
"step": 1870
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 0.8888362171130599,
|
| 2631 |
+
"grad_norm": 0.05263051860972721,
|
| 2632 |
+
"learning_rate": 1.1111283420922528e-05,
|
| 2633 |
+
"loss": 0.5788,
|
| 2634 |
+
"step": 1875
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 0.8912064470253614,
|
| 2638 |
+
"grad_norm": 0.05226957394199282,
|
| 2639 |
+
"learning_rate": 1.0647155102595406e-05,
|
| 2640 |
+
"loss": 0.5665,
|
| 2641 |
+
"step": 1880
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.893576676937663,
|
| 2645 |
+
"grad_norm": 0.05393679935274463,
|
| 2646 |
+
"learning_rate": 1.0192571449449567e-05,
|
| 2647 |
+
"loss": 0.5782,
|
| 2648 |
+
"step": 1885
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 0.8959469068499645,
|
| 2652 |
+
"grad_norm": 0.05290726883653969,
|
| 2653 |
+
"learning_rate": 9.747563597187791e-06,
|
| 2654 |
+
"loss": 0.5725,
|
| 2655 |
+
"step": 1890
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.898317136762266,
|
| 2659 |
+
"grad_norm": 0.052272530064695,
|
| 2660 |
+
"learning_rate": 9.312162025639575e-06,
|
| 2661 |
+
"loss": 0.5696,
|
| 2662 |
+
"step": 1895
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.9006873666745674,
|
| 2666 |
+
"grad_norm": 0.052434933736767136,
|
| 2667 |
+
"learning_rate": 8.886396556673476e-06,
|
| 2668 |
+
"loss": 0.5799,
|
| 2669 |
+
"step": 1900
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.9030575965868689,
|
| 2673 |
+
"grad_norm": 0.049740509342076265,
|
| 2674 |
+
"learning_rate": 8.470296352154466e-06,
|
| 2675 |
+
"loss": 0.5586,
|
| 2676 |
+
"step": 1905
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.9054278264991704,
|
| 2680 |
+
"grad_norm": 0.05205499325333267,
|
| 2681 |
+
"learning_rate": 8.063889911946597e-06,
|
| 2682 |
+
"loss": 0.5715,
|
| 2683 |
+
"step": 1910
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.9077980564114719,
|
| 2687 |
+
"grad_norm": 0.05138219725318278,
|
| 2688 |
+
"learning_rate": 7.667205071960897e-06,
|
| 2689 |
+
"loss": 0.5706,
|
| 2690 |
+
"step": 1915
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.9101682863237734,
|
| 2694 |
+
"grad_norm": 0.053660634763309435,
|
| 2695 |
+
"learning_rate": 7.280269002249006e-06,
|
| 2696 |
+
"loss": 0.5761,
|
| 2697 |
+
"step": 1920
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.9125385162360748,
|
| 2701 |
+
"grad_norm": 0.05161198079700086,
|
| 2702 |
+
"learning_rate": 6.903108205141922e-06,
|
| 2703 |
+
"loss": 0.5676,
|
| 2704 |
+
"step": 1925
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.9149087461483764,
|
| 2708 |
+
"grad_norm": 0.05035126816965241,
|
| 2709 |
+
"learning_rate": 6.535748513435063e-06,
|
| 2710 |
+
"loss": 0.5661,
|
| 2711 |
+
"step": 1930
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.9172789760606779,
|
| 2715 |
+
"grad_norm": 0.05307717275910678,
|
| 2716 |
+
"learning_rate": 6.178215088618754e-06,
|
| 2717 |
+
"loss": 0.5679,
|
| 2718 |
+
"step": 1935
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.9196492059729794,
|
| 2722 |
+
"grad_norm": 0.05151414534729879,
|
| 2723 |
+
"learning_rate": 5.830532419154904e-06,
|
| 2724 |
+
"loss": 0.5655,
|
| 2725 |
+
"step": 1940
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.9220194358852809,
|
| 2729 |
+
"grad_norm": 0.05138066400564488,
|
| 2730 |
+
"learning_rate": 5.492724318799613e-06,
|
| 2731 |
+
"loss": 0.5745,
|
| 2732 |
+
"step": 1945
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.9243896657975824,
|
| 2736 |
+
"grad_norm": 0.053677080886407436,
|
| 2737 |
+
"learning_rate": 5.164813924972294e-06,
|
| 2738 |
+
"loss": 0.5845,
|
| 2739 |
+
"step": 1950
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 0.9267598957098838,
|
| 2743 |
+
"grad_norm": 0.052774180712557975,
|
| 2744 |
+
"learning_rate": 4.846823697170793e-06,
|
| 2745 |
+
"loss": 0.5766,
|
| 2746 |
+
"step": 1955
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 0.9291301256221853,
|
| 2750 |
+
"grad_norm": 0.0521259881298222,
|
| 2751 |
+
"learning_rate": 4.538775415433099e-06,
|
| 2752 |
+
"loss": 0.5817,
|
| 2753 |
+
"step": 1960
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 0.9315003555344868,
|
| 2757 |
+
"grad_norm": 0.05114820813668919,
|
| 2758 |
+
"learning_rate": 4.240690178845507e-06,
|
| 2759 |
+
"loss": 0.586,
|
| 2760 |
+
"step": 1965
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 0.9338705854467884,
|
| 2764 |
+
"grad_norm": 0.054406572055921386,
|
| 2765 |
+
"learning_rate": 3.952588404097706e-06,
|
| 2766 |
+
"loss": 0.5763,
|
| 2767 |
+
"step": 1970
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 0.9362408153590899,
|
| 2771 |
+
"grad_norm": 0.051364642860872965,
|
| 2772 |
+
"learning_rate": 3.6744898240840816e-06,
|
| 2773 |
+
"loss": 0.5724,
|
| 2774 |
+
"step": 1975
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 0.9386110452713913,
|
| 2778 |
+
"grad_norm": 0.051346137904923476,
|
| 2779 |
+
"learning_rate": 3.4064134865523617e-06,
|
| 2780 |
+
"loss": 0.5682,
|
| 2781 |
+
"step": 1980
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 0.9409812751836928,
|
| 2785 |
+
"grad_norm": 0.051110496158009784,
|
| 2786 |
+
"learning_rate": 3.148377752798914e-06,
|
| 2787 |
+
"loss": 0.5693,
|
| 2788 |
+
"step": 1985
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 0.9433515050959943,
|
| 2792 |
+
"grad_norm": 0.05315034654966591,
|
| 2793 |
+
"learning_rate": 2.9004002964111485e-06,
|
| 2794 |
+
"loss": 0.5755,
|
| 2795 |
+
"step": 1990
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 0.9457217350082958,
|
| 2799 |
+
"grad_norm": 0.052170597166830944,
|
| 2800 |
+
"learning_rate": 2.66249810205697e-06,
|
| 2801 |
+
"loss": 0.5796,
|
| 2802 |
+
"step": 1995
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 0.9480919649205973,
|
| 2806 |
+
"grad_norm": 0.05200440928257493,
|
| 2807 |
+
"learning_rate": 2.4346874643214774e-06,
|
| 2808 |
+
"loss": 0.5717,
|
| 2809 |
+
"step": 2000
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 0.9504621948328988,
|
| 2813 |
+
"grad_norm": 0.05163936837645852,
|
| 2814 |
+
"learning_rate": 2.2169839865909036e-06,
|
| 2815 |
+
"loss": 0.5631,
|
| 2816 |
+
"step": 2005
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 0.9528324247452002,
|
| 2820 |
+
"grad_norm": 0.049675359269013855,
|
| 2821 |
+
"learning_rate": 2.009402579983921e-06,
|
| 2822 |
+
"loss": 0.5716,
|
| 2823 |
+
"step": 2010
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 0.9552026546575018,
|
| 2827 |
+
"grad_norm": 0.051088496480251495,
|
| 2828 |
+
"learning_rate": 1.8119574623302436e-06,
|
| 2829 |
+
"loss": 0.561,
|
| 2830 |
+
"step": 2015
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 0.9575728845698033,
|
| 2834 |
+
"grad_norm": 0.05018933256487338,
|
| 2835 |
+
"learning_rate": 1.6246621571969376e-06,
|
| 2836 |
+
"loss": 0.5686,
|
| 2837 |
+
"step": 2020
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 0.9599431144821048,
|
| 2841 |
+
"grad_norm": 0.05409872852152585,
|
| 2842 |
+
"learning_rate": 1.447529492962063e-06,
|
| 2843 |
+
"loss": 0.5658,
|
| 2844 |
+
"step": 2025
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 0.9623133443944063,
|
| 2848 |
+
"grad_norm": 0.05217850594057917,
|
| 2849 |
+
"learning_rate": 1.280571601936009e-06,
|
| 2850 |
+
"loss": 0.5643,
|
| 2851 |
+
"step": 2030
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 0.9646835743067077,
|
| 2855 |
+
"grad_norm": 0.052394586476799676,
|
| 2856 |
+
"learning_rate": 1.1237999195306102e-06,
|
| 2857 |
+
"loss": 0.575,
|
| 2858 |
+
"step": 2035
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 0.9670538042190092,
|
| 2862 |
+
"grad_norm": 0.05288849803820924,
|
| 2863 |
+
"learning_rate": 9.77225183475855e-07,
|
| 2864 |
+
"loss": 0.5753,
|
| 2865 |
+
"step": 2040
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 0.9694240341313107,
|
| 2869 |
+
"grad_norm": 0.05313372067559112,
|
| 2870 |
+
"learning_rate": 8.408574330843576e-07,
|
| 2871 |
+
"loss": 0.5745,
|
| 2872 |
+
"step": 2045
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 0.9717942640436122,
|
| 2876 |
+
"grad_norm": 0.0529192880219954,
|
| 2877 |
+
"learning_rate": 7.147060085638757e-07,
|
| 2878 |
+
"loss": 0.5749,
|
| 2879 |
+
"step": 2050
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 0.9741644939559138,
|
| 2883 |
+
"grad_norm": 0.05401698190339256,
|
| 2884 |
+
"learning_rate": 5.987795503774706e-07,
|
| 2885 |
+
"loss": 0.5743,
|
| 2886 |
+
"step": 2055
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 0.9765347238682153,
|
| 2890 |
+
"grad_norm": 0.05269598387655656,
|
| 2891 |
+
"learning_rate": 4.930859986517322e-07,
|
| 2892 |
+
"loss": 0.5696,
|
| 2893 |
+
"step": 2060
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 0.9789049537805167,
|
| 2897 |
+
"grad_norm": 0.05202660809884899,
|
| 2898 |
+
"learning_rate": 3.9763259263292936e-07,
|
| 2899 |
+
"loss": 0.5596,
|
| 2900 |
+
"step": 2065
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 0.9812751836928182,
|
| 2904 |
+
"grad_norm": 0.06891661422461348,
|
| 2905 |
+
"learning_rate": 3.1242587019117417e-07,
|
| 2906 |
+
"loss": 0.5855,
|
| 2907 |
+
"step": 2070
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 0.9836454136051197,
|
| 2911 |
+
"grad_norm": 0.04973359674018789,
|
| 2912 |
+
"learning_rate": 2.3747166737264622e-07,
|
| 2913 |
+
"loss": 0.5651,
|
| 2914 |
+
"step": 2075
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 0.9860156435174212,
|
| 2918 |
+
"grad_norm": 0.051060019252012756,
|
| 2919 |
+
"learning_rate": 1.7277511799984601e-07,
|
| 2920 |
+
"loss": 0.564,
|
| 2921 |
+
"step": 2080
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 0.9883858734297227,
|
| 2925 |
+
"grad_norm": 0.05056552339235572,
|
| 2926 |
+
"learning_rate": 1.1834065331995935e-07,
|
| 2927 |
+
"loss": 0.5585,
|
| 2928 |
+
"step": 2085
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 0.9907561033420241,
|
| 2932 |
+
"grad_norm": 0.05126255919954496,
|
| 2933 |
+
"learning_rate": 7.41720017013503e-08,
|
| 2934 |
+
"loss": 0.5649,
|
| 2935 |
+
"step": 2090
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 0.9931263332543256,
|
| 2939 |
+
"grad_norm": 0.05801752774330538,
|
| 2940 |
+
"learning_rate": 4.027218837821533e-08,
|
| 2941 |
+
"loss": 0.5712,
|
| 2942 |
+
"step": 2095
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 0.9954965631666272,
|
| 2946 |
+
"grad_norm": 0.05065216789680233,
|
| 2947 |
+
"learning_rate": 1.6643535243365815e-08,
|
| 2948 |
+
"loss": 0.5745,
|
| 2949 |
+
"step": 2100
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 0.9978667930789287,
|
| 2953 |
+
"grad_norm": 0.05241312893170103,
|
| 2954 |
+
"learning_rate": 3.287660689171856e-09,
|
| 2955 |
+
"loss": 0.5783,
|
| 2956 |
+
"step": 2105
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 0.9997629770087698,
|
| 2960 |
+
"eval_loss": 0.5719642639160156,
|
| 2961 |
+
"eval_runtime": 101.3361,
|
| 2962 |
+
"eval_samples_per_second": 296.045,
|
| 2963 |
+
"eval_steps_per_second": 9.256,
|
| 2964 |
+
"step": 2109
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 0.9997629770087698,
|
| 2968 |
+
"step": 2109,
|
| 2969 |
+
"total_flos": 141388749045760.0,
|
| 2970 |
+
"train_loss": 0.6815844999777304,
|
| 2971 |
+
"train_runtime": 3835.2311,
|
| 2972 |
+
"train_samples_per_second": 70.4,
|
| 2973 |
+
"train_steps_per_second": 0.55
|
| 2974 |
+
}
|
| 2975 |
+
],
|
| 2976 |
+
"logging_steps": 5,
|
| 2977 |
+
"max_steps": 2109,
|
| 2978 |
+
"num_input_tokens_seen": 0,
|
| 2979 |
+
"num_train_epochs": 1,
|
| 2980 |
+
"save_steps": 500,
|
| 2981 |
+
"stateful_callbacks": {
|
| 2982 |
+
"TrainerControl": {
|
| 2983 |
+
"args": {
|
| 2984 |
+
"should_epoch_stop": false,
|
| 2985 |
+
"should_evaluate": false,
|
| 2986 |
+
"should_log": false,
|
| 2987 |
+
"should_save": false,
|
| 2988 |
+
"should_training_stop": false
|
| 2989 |
+
},
|
| 2990 |
+
"attributes": {}
|
| 2991 |
+
}
|
| 2992 |
+
},
|
| 2993 |
+
"total_flos": 141388749045760.0,
|
| 2994 |
+
"train_batch_size": 4,
|
| 2995 |
+
"trial_name": null,
|
| 2996 |
+
"trial_params": null
|
| 2997 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fff5cc57f17c9358873ed2440835bcaae8407e38fd23581928961cf8dc37f042
|
| 3 |
+
size 7160
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|