Model save
Browse files- README.md +56 -0
- all_results.json +9 -0
- generation_config.json +7 -0
- model.safetensors +1 -1
- train_results.json +9 -0
- trainer_state.json +2857 -0
README.md
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
model_name: tinyllama-1b-sigmoid-sft-full
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
- trl
|
| 7 |
+
- sft
|
| 8 |
+
licence: license
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# Model Card for tinyllama-1b-sigmoid-sft-full
|
| 12 |
+
|
| 13 |
+
This model is a fine-tuned version of [None](https://huggingface.co/None).
|
| 14 |
+
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 15 |
+
|
| 16 |
+
## Quick start
|
| 17 |
+
|
| 18 |
+
```python
|
| 19 |
+
from transformers import pipeline
|
| 20 |
+
|
| 21 |
+
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?"
|
| 22 |
+
generator = pipeline("text-generation", model="BrianGu/tinyllama-1b-sigmoid-sft-full", device="cuda")
|
| 23 |
+
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
|
| 24 |
+
print(output["generated_text"])
|
| 25 |
+
```
|
| 26 |
+
|
| 27 |
+
## Training procedure
|
| 28 |
+
|
| 29 |
+
[<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/braingu/sft/runs/lnwljtwq)
|
| 30 |
+
|
| 31 |
+
This model was trained with SFT.
|
| 32 |
+
|
| 33 |
+
### Framework versions
|
| 34 |
+
|
| 35 |
+
- TRL: 0.12.1
|
| 36 |
+
- Transformers: 4.46.2
|
| 37 |
+
- Pytorch: 2.5.1
|
| 38 |
+
- Datasets: 3.1.0
|
| 39 |
+
- Tokenizers: 0.20.3
|
| 40 |
+
|
| 41 |
+
## Citations
|
| 42 |
+
|
| 43 |
+
|
| 44 |
+
|
| 45 |
+
Cite TRL as:
|
| 46 |
+
|
| 47 |
+
```bibtex
|
| 48 |
+
@misc{vonwerra2022trl,
|
| 49 |
+
title = {{TRL: Transformer Reinforcement Learning}},
|
| 50 |
+
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},
|
| 51 |
+
year = 2020,
|
| 52 |
+
journal = {GitHub repository},
|
| 53 |
+
publisher = {GitHub},
|
| 54 |
+
howpublished = {\url{https://github.com/huggingface/trl}}
|
| 55 |
+
}
|
| 56 |
+
```
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 145100101386240.0,
|
| 4 |
+
"train_loss": 1.7762367073948924,
|
| 5 |
+
"train_runtime": 4136.0894,
|
| 6 |
+
"train_samples": 207864,
|
| 7 |
+
"train_samples_per_second": 30.978,
|
| 8 |
+
"train_steps_per_second": 0.484
|
| 9 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"transformers_version": "4.46.2",
|
| 6 |
+
"use_cache": false
|
| 7 |
+
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2674076304
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7030065de5b9bc32f42f2be0d88c70e14f0efe432c5637f0015b163f4adb49bb
|
| 3 |
size 2674076304
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 145100101386240.0,
|
| 4 |
+
"train_loss": 1.7762367073948924,
|
| 5 |
+
"train_runtime": 4136.0894,
|
| 6 |
+
"train_samples": 207864,
|
| 7 |
+
"train_samples_per_second": 30.978,
|
| 8 |
+
"train_steps_per_second": 0.484
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2857 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2002,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0004995004995004995,
|
| 13 |
+
"grad_norm": 1.1599862583346459,
|
| 14 |
+
"learning_rate": 9.950248756218906e-08,
|
| 15 |
+
"loss": 2.4381,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.0024975024975024975,
|
| 20 |
+
"grad_norm": 1.3646452392145616,
|
| 21 |
+
"learning_rate": 4.975124378109453e-07,
|
| 22 |
+
"loss": 2.3925,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.004995004995004995,
|
| 27 |
+
"grad_norm": 1.050547128080392,
|
| 28 |
+
"learning_rate": 9.950248756218907e-07,
|
| 29 |
+
"loss": 2.3899,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.007492507492507493,
|
| 34 |
+
"grad_norm": 1.1106843586026567,
|
| 35 |
+
"learning_rate": 1.4925373134328358e-06,
|
| 36 |
+
"loss": 2.3606,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.00999000999000999,
|
| 41 |
+
"grad_norm": 1.0868632779387772,
|
| 42 |
+
"learning_rate": 1.9900497512437813e-06,
|
| 43 |
+
"loss": 2.3782,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.012487512487512488,
|
| 48 |
+
"grad_norm": 0.8822872185142195,
|
| 49 |
+
"learning_rate": 2.4875621890547264e-06,
|
| 50 |
+
"loss": 2.3681,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.014985014985014986,
|
| 55 |
+
"grad_norm": 0.6990472283962801,
|
| 56 |
+
"learning_rate": 2.9850746268656716e-06,
|
| 57 |
+
"loss": 2.3332,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.017482517482517484,
|
| 62 |
+
"grad_norm": 0.6640215974826378,
|
| 63 |
+
"learning_rate": 3.4825870646766175e-06,
|
| 64 |
+
"loss": 2.3265,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.01998001998001998,
|
| 69 |
+
"grad_norm": 0.6451421174329706,
|
| 70 |
+
"learning_rate": 3.980099502487563e-06,
|
| 71 |
+
"loss": 2.2609,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.022477522477522476,
|
| 76 |
+
"grad_norm": 0.4613452178440462,
|
| 77 |
+
"learning_rate": 4.477611940298508e-06,
|
| 78 |
+
"loss": 2.1835,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.024975024975024976,
|
| 83 |
+
"grad_norm": 0.36216302524794514,
|
| 84 |
+
"learning_rate": 4.975124378109453e-06,
|
| 85 |
+
"loss": 2.1793,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.027472527472527472,
|
| 90 |
+
"grad_norm": 0.29388450685747625,
|
| 91 |
+
"learning_rate": 5.472636815920398e-06,
|
| 92 |
+
"loss": 2.1195,
|
| 93 |
+
"step": 55
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.029970029970029972,
|
| 97 |
+
"grad_norm": 0.25381832457288017,
|
| 98 |
+
"learning_rate": 5.970149253731343e-06,
|
| 99 |
+
"loss": 2.065,
|
| 100 |
+
"step": 60
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.032467532467532464,
|
| 104 |
+
"grad_norm": 0.23481350483234162,
|
| 105 |
+
"learning_rate": 6.46766169154229e-06,
|
| 106 |
+
"loss": 2.1047,
|
| 107 |
+
"step": 65
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.03496503496503497,
|
| 111 |
+
"grad_norm": 0.20628569151997012,
|
| 112 |
+
"learning_rate": 6.965174129353235e-06,
|
| 113 |
+
"loss": 2.0828,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.037462537462537464,
|
| 118 |
+
"grad_norm": 0.19709869224218293,
|
| 119 |
+
"learning_rate": 7.46268656716418e-06,
|
| 120 |
+
"loss": 2.0265,
|
| 121 |
+
"step": 75
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.03996003996003996,
|
| 125 |
+
"grad_norm": 0.2006237162245652,
|
| 126 |
+
"learning_rate": 7.960199004975125e-06,
|
| 127 |
+
"loss": 2.0299,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.042457542457542456,
|
| 132 |
+
"grad_norm": 0.18543831140294761,
|
| 133 |
+
"learning_rate": 8.45771144278607e-06,
|
| 134 |
+
"loss": 2.0171,
|
| 135 |
+
"step": 85
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.04495504495504495,
|
| 139 |
+
"grad_norm": 0.18480210061309715,
|
| 140 |
+
"learning_rate": 8.955223880597016e-06,
|
| 141 |
+
"loss": 2.0388,
|
| 142 |
+
"step": 90
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.047452547452547456,
|
| 146 |
+
"grad_norm": 0.18002159323152245,
|
| 147 |
+
"learning_rate": 9.45273631840796e-06,
|
| 148 |
+
"loss": 2.0161,
|
| 149 |
+
"step": 95
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.04995004995004995,
|
| 153 |
+
"grad_norm": 0.18038747124195287,
|
| 154 |
+
"learning_rate": 9.950248756218906e-06,
|
| 155 |
+
"loss": 1.9666,
|
| 156 |
+
"step": 100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.05244755244755245,
|
| 160 |
+
"grad_norm": 0.18508558771562467,
|
| 161 |
+
"learning_rate": 1.0447761194029851e-05,
|
| 162 |
+
"loss": 1.9788,
|
| 163 |
+
"step": 105
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.054945054945054944,
|
| 167 |
+
"grad_norm": 0.1753326294106231,
|
| 168 |
+
"learning_rate": 1.0945273631840796e-05,
|
| 169 |
+
"loss": 1.9935,
|
| 170 |
+
"step": 110
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.05744255744255744,
|
| 174 |
+
"grad_norm": 0.17338505913330995,
|
| 175 |
+
"learning_rate": 1.1442786069651741e-05,
|
| 176 |
+
"loss": 1.982,
|
| 177 |
+
"step": 115
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.059940059940059943,
|
| 181 |
+
"grad_norm": 0.17069633894083674,
|
| 182 |
+
"learning_rate": 1.1940298507462686e-05,
|
| 183 |
+
"loss": 1.9677,
|
| 184 |
+
"step": 120
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.06243756243756244,
|
| 188 |
+
"grad_norm": 0.17033240645669973,
|
| 189 |
+
"learning_rate": 1.2437810945273631e-05,
|
| 190 |
+
"loss": 1.9579,
|
| 191 |
+
"step": 125
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.06493506493506493,
|
| 195 |
+
"grad_norm": 0.16801885256509916,
|
| 196 |
+
"learning_rate": 1.293532338308458e-05,
|
| 197 |
+
"loss": 1.9328,
|
| 198 |
+
"step": 130
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.06743256743256744,
|
| 202 |
+
"grad_norm": 0.16854179300669697,
|
| 203 |
+
"learning_rate": 1.3432835820895525e-05,
|
| 204 |
+
"loss": 1.9424,
|
| 205 |
+
"step": 135
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.06993006993006994,
|
| 209 |
+
"grad_norm": 0.16787976921679182,
|
| 210 |
+
"learning_rate": 1.393034825870647e-05,
|
| 211 |
+
"loss": 1.9229,
|
| 212 |
+
"step": 140
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.07242757242757243,
|
| 216 |
+
"grad_norm": 0.167368866372539,
|
| 217 |
+
"learning_rate": 1.4427860696517415e-05,
|
| 218 |
+
"loss": 1.9537,
|
| 219 |
+
"step": 145
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.07492507492507493,
|
| 223 |
+
"grad_norm": 0.16524535846346908,
|
| 224 |
+
"learning_rate": 1.492537313432836e-05,
|
| 225 |
+
"loss": 1.911,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.07742257742257742,
|
| 230 |
+
"grad_norm": 0.16714889916639722,
|
| 231 |
+
"learning_rate": 1.5422885572139307e-05,
|
| 232 |
+
"loss": 1.9399,
|
| 233 |
+
"step": 155
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.07992007992007992,
|
| 237 |
+
"grad_norm": 0.16617978712852743,
|
| 238 |
+
"learning_rate": 1.592039800995025e-05,
|
| 239 |
+
"loss": 1.9203,
|
| 240 |
+
"step": 160
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.08241758241758242,
|
| 244 |
+
"grad_norm": 0.16512217331592868,
|
| 245 |
+
"learning_rate": 1.6417910447761197e-05,
|
| 246 |
+
"loss": 1.923,
|
| 247 |
+
"step": 165
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.08491508491508491,
|
| 251 |
+
"grad_norm": 0.16496696956998044,
|
| 252 |
+
"learning_rate": 1.691542288557214e-05,
|
| 253 |
+
"loss": 1.9442,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.08741258741258741,
|
| 258 |
+
"grad_norm": 0.16376024451930163,
|
| 259 |
+
"learning_rate": 1.7412935323383088e-05,
|
| 260 |
+
"loss": 1.931,
|
| 261 |
+
"step": 175
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.0899100899100899,
|
| 265 |
+
"grad_norm": 0.1640420678366522,
|
| 266 |
+
"learning_rate": 1.791044776119403e-05,
|
| 267 |
+
"loss": 1.908,
|
| 268 |
+
"step": 180
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.0924075924075924,
|
| 272 |
+
"grad_norm": 0.160666099150629,
|
| 273 |
+
"learning_rate": 1.8407960199004978e-05,
|
| 274 |
+
"loss": 1.9139,
|
| 275 |
+
"step": 185
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.09490509490509491,
|
| 279 |
+
"grad_norm": 0.16238948145763205,
|
| 280 |
+
"learning_rate": 1.890547263681592e-05,
|
| 281 |
+
"loss": 1.8926,
|
| 282 |
+
"step": 190
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.09740259740259741,
|
| 286 |
+
"grad_norm": 0.163531659071721,
|
| 287 |
+
"learning_rate": 1.9402985074626868e-05,
|
| 288 |
+
"loss": 1.899,
|
| 289 |
+
"step": 195
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.0999000999000999,
|
| 293 |
+
"grad_norm": 0.16286090035138232,
|
| 294 |
+
"learning_rate": 1.990049751243781e-05,
|
| 295 |
+
"loss": 1.9091,
|
| 296 |
+
"step": 200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.1023976023976024,
|
| 300 |
+
"grad_norm": 0.16060757439288406,
|
| 301 |
+
"learning_rate": 1.9999756577597317e-05,
|
| 302 |
+
"loss": 1.8851,
|
| 303 |
+
"step": 205
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.1048951048951049,
|
| 307 |
+
"grad_norm": 0.16221564755663112,
|
| 308 |
+
"learning_rate": 1.9998767694397236e-05,
|
| 309 |
+
"loss": 1.8779,
|
| 310 |
+
"step": 210
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.10739260739260739,
|
| 314 |
+
"grad_norm": 0.158639070485282,
|
| 315 |
+
"learning_rate": 1.99970182116654e-05,
|
| 316 |
+
"loss": 1.8574,
|
| 317 |
+
"step": 215
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.10989010989010989,
|
| 321 |
+
"grad_norm": 0.1602764625642043,
|
| 322 |
+
"learning_rate": 1.9994508262483786e-05,
|
| 323 |
+
"loss": 1.8777,
|
| 324 |
+
"step": 220
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.11238761238761238,
|
| 328 |
+
"grad_norm": 0.16307835969521423,
|
| 329 |
+
"learning_rate": 1.999123803778254e-05,
|
| 330 |
+
"loss": 1.9083,
|
| 331 |
+
"step": 225
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.11488511488511488,
|
| 335 |
+
"grad_norm": 0.1628647640219108,
|
| 336 |
+
"learning_rate": 1.998720778632546e-05,
|
| 337 |
+
"loss": 1.8748,
|
| 338 |
+
"step": 230
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.11738261738261738,
|
| 342 |
+
"grad_norm": 0.16061369451851806,
|
| 343 |
+
"learning_rate": 1.9982417814691048e-05,
|
| 344 |
+
"loss": 1.8514,
|
| 345 |
+
"step": 235
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.11988011988011989,
|
| 349 |
+
"grad_norm": 0.1623492271600646,
|
| 350 |
+
"learning_rate": 1.997686848724924e-05,
|
| 351 |
+
"loss": 1.8334,
|
| 352 |
+
"step": 240
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.12237762237762238,
|
| 356 |
+
"grad_norm": 0.16052653488024277,
|
| 357 |
+
"learning_rate": 1.997056022613363e-05,
|
| 358 |
+
"loss": 1.8828,
|
| 359 |
+
"step": 245
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.12487512487512488,
|
| 363 |
+
"grad_norm": 0.16186662472616486,
|
| 364 |
+
"learning_rate": 1.9963493511209405e-05,
|
| 365 |
+
"loss": 1.8689,
|
| 366 |
+
"step": 250
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.12737262737262736,
|
| 370 |
+
"grad_norm": 0.15814062813935195,
|
| 371 |
+
"learning_rate": 1.9955668880036812e-05,
|
| 372 |
+
"loss": 1.8633,
|
| 373 |
+
"step": 255
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.12987012987012986,
|
| 377 |
+
"grad_norm": 0.16042937756824743,
|
| 378 |
+
"learning_rate": 1.9947086927830277e-05,
|
| 379 |
+
"loss": 1.859,
|
| 380 |
+
"step": 260
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.13236763236763235,
|
| 384 |
+
"grad_norm": 0.15749320226952399,
|
| 385 |
+
"learning_rate": 1.9937748307413134e-05,
|
| 386 |
+
"loss": 1.8324,
|
| 387 |
+
"step": 265
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.13486513486513488,
|
| 391 |
+
"grad_norm": 0.15932775031704177,
|
| 392 |
+
"learning_rate": 1.9927653729167957e-05,
|
| 393 |
+
"loss": 1.8403,
|
| 394 |
+
"step": 270
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.13736263736263737,
|
| 398 |
+
"grad_norm": 0.15539081264737326,
|
| 399 |
+
"learning_rate": 1.9916803960982518e-05,
|
| 400 |
+
"loss": 1.8352,
|
| 401 |
+
"step": 275
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.13986013986013987,
|
| 405 |
+
"grad_norm": 0.1616390225384804,
|
| 406 |
+
"learning_rate": 1.9905199828191385e-05,
|
| 407 |
+
"loss": 1.8703,
|
| 408 |
+
"step": 280
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.14235764235764237,
|
| 412 |
+
"grad_norm": 0.1608186724233932,
|
| 413 |
+
"learning_rate": 1.9892842213513135e-05,
|
| 414 |
+
"loss": 1.8503,
|
| 415 |
+
"step": 285
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.14485514485514486,
|
| 419 |
+
"grad_norm": 0.15963816595377592,
|
| 420 |
+
"learning_rate": 1.98797320569832e-05,
|
| 421 |
+
"loss": 1.8459,
|
| 422 |
+
"step": 290
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.14735264735264736,
|
| 426 |
+
"grad_norm": 0.1582839014078666,
|
| 427 |
+
"learning_rate": 1.986587035588237e-05,
|
| 428 |
+
"loss": 1.8553,
|
| 429 |
+
"step": 295
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.14985014985014986,
|
| 433 |
+
"grad_norm": 0.15947984158782932,
|
| 434 |
+
"learning_rate": 1.985125816466092e-05,
|
| 435 |
+
"loss": 1.8418,
|
| 436 |
+
"step": 300
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.15234765234765235,
|
| 440 |
+
"grad_norm": 0.15926277299896913,
|
| 441 |
+
"learning_rate": 1.9835896594858405e-05,
|
| 442 |
+
"loss": 1.8398,
|
| 443 |
+
"step": 305
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.15484515484515485,
|
| 447 |
+
"grad_norm": 0.1619303456318674,
|
| 448 |
+
"learning_rate": 1.9819786815019108e-05,
|
| 449 |
+
"loss": 1.8453,
|
| 450 |
+
"step": 310
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.15734265734265734,
|
| 454 |
+
"grad_norm": 0.15663126843033132,
|
| 455 |
+
"learning_rate": 1.9802930050603143e-05,
|
| 456 |
+
"loss": 1.7968,
|
| 457 |
+
"step": 315
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.15984015984015984,
|
| 461 |
+
"grad_norm": 0.15428142491915428,
|
| 462 |
+
"learning_rate": 1.9785327583893233e-05,
|
| 463 |
+
"loss": 1.7914,
|
| 464 |
+
"step": 320
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.16233766233766234,
|
| 468 |
+
"grad_norm": 0.1535269165426423,
|
| 469 |
+
"learning_rate": 1.9766980753897186e-05,
|
| 470 |
+
"loss": 1.8398,
|
| 471 |
+
"step": 325
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.16483516483516483,
|
| 475 |
+
"grad_norm": 0.1566922978711067,
|
| 476 |
+
"learning_rate": 1.974789095624601e-05,
|
| 477 |
+
"loss": 1.8197,
|
| 478 |
+
"step": 330
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.16733266733266733,
|
| 482 |
+
"grad_norm": 0.16077008857885935,
|
| 483 |
+
"learning_rate": 1.972805964308778e-05,
|
| 484 |
+
"loss": 1.827,
|
| 485 |
+
"step": 335
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.16983016983016982,
|
| 489 |
+
"grad_norm": 0.1595092032589525,
|
| 490 |
+
"learning_rate": 1.9707488322977137e-05,
|
| 491 |
+
"loss": 1.8202,
|
| 492 |
+
"step": 340
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.17232767232767232,
|
| 496 |
+
"grad_norm": 0.16187621193128984,
|
| 497 |
+
"learning_rate": 1.968617856076056e-05,
|
| 498 |
+
"loss": 1.8462,
|
| 499 |
+
"step": 345
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.17482517482517482,
|
| 503 |
+
"grad_norm": 0.15740467422342347,
|
| 504 |
+
"learning_rate": 1.9664131977457334e-05,
|
| 505 |
+
"loss": 1.8285,
|
| 506 |
+
"step": 350
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.1773226773226773,
|
| 510 |
+
"grad_norm": 0.15653103603511576,
|
| 511 |
+
"learning_rate": 1.964135025013621e-05,
|
| 512 |
+
"loss": 1.8524,
|
| 513 |
+
"step": 355
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.1798201798201798,
|
| 517 |
+
"grad_norm": 0.16162833026081047,
|
| 518 |
+
"learning_rate": 1.9617835111787867e-05,
|
| 519 |
+
"loss": 1.8384,
|
| 520 |
+
"step": 360
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.1823176823176823,
|
| 524 |
+
"grad_norm": 0.15473723241970383,
|
| 525 |
+
"learning_rate": 1.9593588351193052e-05,
|
| 526 |
+
"loss": 1.8139,
|
| 527 |
+
"step": 365
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.1848151848151848,
|
| 531 |
+
"grad_norm": 0.15885275250786435,
|
| 532 |
+
"learning_rate": 1.9568611812786532e-05,
|
| 533 |
+
"loss": 1.8218,
|
| 534 |
+
"step": 370
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.18731268731268733,
|
| 538 |
+
"grad_norm": 0.1568218347795679,
|
| 539 |
+
"learning_rate": 1.9542907396516785e-05,
|
| 540 |
+
"loss": 1.8099,
|
| 541 |
+
"step": 375
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.18981018981018982,
|
| 545 |
+
"grad_norm": 0.15705793055606954,
|
| 546 |
+
"learning_rate": 1.9516477057701466e-05,
|
| 547 |
+
"loss": 1.8264,
|
| 548 |
+
"step": 380
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.19230769230769232,
|
| 552 |
+
"grad_norm": 0.15732378505813635,
|
| 553 |
+
"learning_rate": 1.9489322806878656e-05,
|
| 554 |
+
"loss": 1.7763,
|
| 555 |
+
"step": 385
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.19480519480519481,
|
| 559 |
+
"grad_norm": 0.16123492252157817,
|
| 560 |
+
"learning_rate": 1.9461446709653957e-05,
|
| 561 |
+
"loss": 1.8205,
|
| 562 |
+
"step": 390
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.1973026973026973,
|
| 566 |
+
"grad_norm": 0.15737920869870928,
|
| 567 |
+
"learning_rate": 1.9432850886543326e-05,
|
| 568 |
+
"loss": 1.8145,
|
| 569 |
+
"step": 395
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.1998001998001998,
|
| 573 |
+
"grad_norm": 0.15899572770709094,
|
| 574 |
+
"learning_rate": 1.9403537512811787e-05,
|
| 575 |
+
"loss": 1.8273,
|
| 576 |
+
"step": 400
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.2022977022977023,
|
| 580 |
+
"grad_norm": 0.15657272543340842,
|
| 581 |
+
"learning_rate": 1.9373508818307968e-05,
|
| 582 |
+
"loss": 1.7994,
|
| 583 |
+
"step": 405
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.2047952047952048,
|
| 587 |
+
"grad_norm": 0.15774961567754203,
|
| 588 |
+
"learning_rate": 1.934276708729445e-05,
|
| 589 |
+
"loss": 1.8152,
|
| 590 |
+
"step": 410
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.2072927072927073,
|
| 594 |
+
"grad_norm": 0.1597456026863612,
|
| 595 |
+
"learning_rate": 1.931131465827403e-05,
|
| 596 |
+
"loss": 1.8067,
|
| 597 |
+
"step": 415
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.2097902097902098,
|
| 601 |
+
"grad_norm": 0.15874324914863305,
|
| 602 |
+
"learning_rate": 1.927915392381183e-05,
|
| 603 |
+
"loss": 1.8113,
|
| 604 |
+
"step": 420
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.2122877122877123,
|
| 608 |
+
"grad_norm": 0.15739095253831112,
|
| 609 |
+
"learning_rate": 1.924628733035327e-05,
|
| 610 |
+
"loss": 1.7572,
|
| 611 |
+
"step": 425
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.21478521478521478,
|
| 615 |
+
"grad_norm": 0.15745037871244488,
|
| 616 |
+
"learning_rate": 1.921271737803802e-05,
|
| 617 |
+
"loss": 1.7815,
|
| 618 |
+
"step": 430
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.21728271728271728,
|
| 622 |
+
"grad_norm": 0.1574617354946401,
|
| 623 |
+
"learning_rate": 1.9178446620509762e-05,
|
| 624 |
+
"loss": 1.8072,
|
| 625 |
+
"step": 435
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.21978021978021978,
|
| 629 |
+
"grad_norm": 0.16217729499687356,
|
| 630 |
+
"learning_rate": 1.9143477664721958e-05,
|
| 631 |
+
"loss": 1.8034,
|
| 632 |
+
"step": 440
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.22227772227772227,
|
| 636 |
+
"grad_norm": 0.16044737049383317,
|
| 637 |
+
"learning_rate": 1.910781317073956e-05,
|
| 638 |
+
"loss": 1.8369,
|
| 639 |
+
"step": 445
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.22477522477522477,
|
| 643 |
+
"grad_norm": 0.1573797424088386,
|
| 644 |
+
"learning_rate": 1.9071455851536624e-05,
|
| 645 |
+
"loss": 1.7681,
|
| 646 |
+
"step": 450
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.22727272727272727,
|
| 650 |
+
"grad_norm": 0.15944418619813316,
|
| 651 |
+
"learning_rate": 1.9034408472789966e-05,
|
| 652 |
+
"loss": 1.8018,
|
| 653 |
+
"step": 455
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.22977022977022976,
|
| 657 |
+
"grad_norm": 0.1567957354310453,
|
| 658 |
+
"learning_rate": 1.899667385266876e-05,
|
| 659 |
+
"loss": 1.8262,
|
| 660 |
+
"step": 460
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.23226773226773226,
|
| 664 |
+
"grad_norm": 0.15732076579865928,
|
| 665 |
+
"learning_rate": 1.8958254861620175e-05,
|
| 666 |
+
"loss": 1.8065,
|
| 667 |
+
"step": 465
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.23476523476523475,
|
| 671 |
+
"grad_norm": 0.155336816664159,
|
| 672 |
+
"learning_rate": 1.891915442215101e-05,
|
| 673 |
+
"loss": 1.8053,
|
| 674 |
+
"step": 470
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.23726273726273725,
|
| 678 |
+
"grad_norm": 0.1597669648163907,
|
| 679 |
+
"learning_rate": 1.88793755086054e-05,
|
| 680 |
+
"loss": 1.8183,
|
| 681 |
+
"step": 475
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.23976023976023977,
|
| 685 |
+
"grad_norm": 0.15788615176295445,
|
| 686 |
+
"learning_rate": 1.8838921146938523e-05,
|
| 687 |
+
"loss": 1.8216,
|
| 688 |
+
"step": 480
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.24225774225774227,
|
| 692 |
+
"grad_norm": 0.15554518751652616,
|
| 693 |
+
"learning_rate": 1.8797794414486464e-05,
|
| 694 |
+
"loss": 1.8177,
|
| 695 |
+
"step": 485
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.24475524475524477,
|
| 699 |
+
"grad_norm": 0.15599244982971106,
|
| 700 |
+
"learning_rate": 1.8755998439732092e-05,
|
| 701 |
+
"loss": 1.7595,
|
| 702 |
+
"step": 490
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.24725274725274726,
|
| 706 |
+
"grad_norm": 0.1565896274442155,
|
| 707 |
+
"learning_rate": 1.8713536402067084e-05,
|
| 708 |
+
"loss": 1.7927,
|
| 709 |
+
"step": 495
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.24975024975024976,
|
| 713 |
+
"grad_norm": 0.15811536697664205,
|
| 714 |
+
"learning_rate": 1.8670411531550078e-05,
|
| 715 |
+
"loss": 1.774,
|
| 716 |
+
"step": 500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.25224775224775225,
|
| 720 |
+
"grad_norm": 0.156281445500983,
|
| 721 |
+
"learning_rate": 1.8626627108660967e-05,
|
| 722 |
+
"loss": 1.7896,
|
| 723 |
+
"step": 505
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.2547452547452547,
|
| 727 |
+
"grad_norm": 0.16199506703702074,
|
| 728 |
+
"learning_rate": 1.858218646405134e-05,
|
| 729 |
+
"loss": 1.7957,
|
| 730 |
+
"step": 510
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.25724275724275725,
|
| 734 |
+
"grad_norm": 0.15409864113956373,
|
| 735 |
+
"learning_rate": 1.853709297829113e-05,
|
| 736 |
+
"loss": 1.7681,
|
| 737 |
+
"step": 515
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.2597402597402597,
|
| 741 |
+
"grad_norm": 0.15595897376396248,
|
| 742 |
+
"learning_rate": 1.8491350081611476e-05,
|
| 743 |
+
"loss": 1.7821,
|
| 744 |
+
"step": 520
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.26223776223776224,
|
| 748 |
+
"grad_norm": 0.15981459835452744,
|
| 749 |
+
"learning_rate": 1.8444961253643735e-05,
|
| 750 |
+
"loss": 1.8,
|
| 751 |
+
"step": 525
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.2647352647352647,
|
| 755 |
+
"grad_norm": 0.1598175484696389,
|
| 756 |
+
"learning_rate": 1.839793002315486e-05,
|
| 757 |
+
"loss": 1.798,
|
| 758 |
+
"step": 530
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.26723276723276723,
|
| 762 |
+
"grad_norm": 0.15880648153398907,
|
| 763 |
+
"learning_rate": 1.8350259967778904e-05,
|
| 764 |
+
"loss": 1.8151,
|
| 765 |
+
"step": 535
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.26973026973026976,
|
| 769 |
+
"grad_norm": 0.16237848244779185,
|
| 770 |
+
"learning_rate": 1.8301954713744913e-05,
|
| 771 |
+
"loss": 1.8267,
|
| 772 |
+
"step": 540
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.2722277722277722,
|
| 776 |
+
"grad_norm": 0.15519746128863732,
|
| 777 |
+
"learning_rate": 1.825301793560106e-05,
|
| 778 |
+
"loss": 1.7678,
|
| 779 |
+
"step": 545
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.27472527472527475,
|
| 783 |
+
"grad_norm": 0.15889605012099542,
|
| 784 |
+
"learning_rate": 1.8203453355935138e-05,
|
| 785 |
+
"loss": 1.7552,
|
| 786 |
+
"step": 550
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.2772227772227772,
|
| 790 |
+
"grad_norm": 0.15715544258244915,
|
| 791 |
+
"learning_rate": 1.8153264745091376e-05,
|
| 792 |
+
"loss": 1.7906,
|
| 793 |
+
"step": 555
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.27972027972027974,
|
| 797 |
+
"grad_norm": 0.15630976676944236,
|
| 798 |
+
"learning_rate": 1.8102455920883638e-05,
|
| 799 |
+
"loss": 1.7698,
|
| 800 |
+
"step": 560
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.2822177822177822,
|
| 804 |
+
"grad_norm": 0.15690924431151737,
|
| 805 |
+
"learning_rate": 1.8051030748304995e-05,
|
| 806 |
+
"loss": 1.7676,
|
| 807 |
+
"step": 565
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.28471528471528473,
|
| 811 |
+
"grad_norm": 0.15783500931857442,
|
| 812 |
+
"learning_rate": 1.7998993139233733e-05,
|
| 813 |
+
"loss": 1.768,
|
| 814 |
+
"step": 570
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.2872127872127872,
|
| 818 |
+
"grad_norm": 0.1567537339747436,
|
| 819 |
+
"learning_rate": 1.7946347052135765e-05,
|
| 820 |
+
"loss": 1.766,
|
| 821 |
+
"step": 575
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.2897102897102897,
|
| 825 |
+
"grad_norm": 0.16054044236045276,
|
| 826 |
+
"learning_rate": 1.7893096491763516e-05,
|
| 827 |
+
"loss": 1.7679,
|
| 828 |
+
"step": 580
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.2922077922077922,
|
| 832 |
+
"grad_norm": 0.1559143042109544,
|
| 833 |
+
"learning_rate": 1.783924550885129e-05,
|
| 834 |
+
"loss": 1.7702,
|
| 835 |
+
"step": 585
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.2947052947052947,
|
| 839 |
+
"grad_norm": 0.1576160782660945,
|
| 840 |
+
"learning_rate": 1.7784798199807128e-05,
|
| 841 |
+
"loss": 1.7615,
|
| 842 |
+
"step": 590
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.2972027972027972,
|
| 846 |
+
"grad_norm": 0.15642650102699018,
|
| 847 |
+
"learning_rate": 1.7729758706401198e-05,
|
| 848 |
+
"loss": 1.7765,
|
| 849 |
+
"step": 595
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.2997002997002997,
|
| 853 |
+
"grad_norm": 0.1559306633293477,
|
| 854 |
+
"learning_rate": 1.7674131215450737e-05,
|
| 855 |
+
"loss": 1.771,
|
| 856 |
+
"step": 600
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.3021978021978022,
|
| 860 |
+
"grad_norm": 0.15795527293903494,
|
| 861 |
+
"learning_rate": 1.761791995850156e-05,
|
| 862 |
+
"loss": 1.7941,
|
| 863 |
+
"step": 605
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.3046953046953047,
|
| 867 |
+
"grad_norm": 0.15804826082822157,
|
| 868 |
+
"learning_rate": 1.756112921150616e-05,
|
| 869 |
+
"loss": 1.7846,
|
| 870 |
+
"step": 610
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.30719280719280717,
|
| 874 |
+
"grad_norm": 0.15566746006494162,
|
| 875 |
+
"learning_rate": 1.7503763294498457e-05,
|
| 876 |
+
"loss": 1.7756,
|
| 877 |
+
"step": 615
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.3096903096903097,
|
| 881 |
+
"grad_norm": 0.1575042305833722,
|
| 882 |
+
"learning_rate": 1.744582657126517e-05,
|
| 883 |
+
"loss": 1.7905,
|
| 884 |
+
"step": 620
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.31218781218781216,
|
| 888 |
+
"grad_norm": 0.16032862889097663,
|
| 889 |
+
"learning_rate": 1.7387323449013847e-05,
|
| 890 |
+
"loss": 1.7862,
|
| 891 |
+
"step": 625
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.3146853146853147,
|
| 895 |
+
"grad_norm": 0.15809033799977093,
|
| 896 |
+
"learning_rate": 1.732825837803765e-05,
|
| 897 |
+
"loss": 1.7702,
|
| 898 |
+
"step": 630
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.31718281718281716,
|
| 902 |
+
"grad_norm": 0.15672979108179055,
|
| 903 |
+
"learning_rate": 1.7268635851376785e-05,
|
| 904 |
+
"loss": 1.76,
|
| 905 |
+
"step": 635
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.3196803196803197,
|
| 909 |
+
"grad_norm": 0.1554900829598635,
|
| 910 |
+
"learning_rate": 1.7208460404476745e-05,
|
| 911 |
+
"loss": 1.7411,
|
| 912 |
+
"step": 640
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.3221778221778222,
|
| 916 |
+
"grad_norm": 0.15739315261539163,
|
| 917 |
+
"learning_rate": 1.7147736614843297e-05,
|
| 918 |
+
"loss": 1.7954,
|
| 919 |
+
"step": 645
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.3246753246753247,
|
| 923 |
+
"grad_norm": 0.15852054226953874,
|
| 924 |
+
"learning_rate": 1.7086469101694256e-05,
|
| 925 |
+
"loss": 1.7627,
|
| 926 |
+
"step": 650
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.3271728271728272,
|
| 930 |
+
"grad_norm": 0.1596482505865284,
|
| 931 |
+
"learning_rate": 1.7024662525608138e-05,
|
| 932 |
+
"loss": 1.751,
|
| 933 |
+
"step": 655
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.32967032967032966,
|
| 937 |
+
"grad_norm": 0.15871675517321407,
|
| 938 |
+
"learning_rate": 1.6962321588169598e-05,
|
| 939 |
+
"loss": 1.7434,
|
| 940 |
+
"step": 660
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.3321678321678322,
|
| 944 |
+
"grad_norm": 0.15766920664878184,
|
| 945 |
+
"learning_rate": 1.6899451031611814e-05,
|
| 946 |
+
"loss": 1.7639,
|
| 947 |
+
"step": 665
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.33466533466533466,
|
| 951 |
+
"grad_norm": 0.155044210346668,
|
| 952 |
+
"learning_rate": 1.6836055638455722e-05,
|
| 953 |
+
"loss": 1.7703,
|
| 954 |
+
"step": 670
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.3371628371628372,
|
| 958 |
+
"grad_norm": 0.1554119105318693,
|
| 959 |
+
"learning_rate": 1.677214023114623e-05,
|
| 960 |
+
"loss": 1.75,
|
| 961 |
+
"step": 675
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.33966033966033965,
|
| 965 |
+
"grad_norm": 0.15802877535499618,
|
| 966 |
+
"learning_rate": 1.670770967168537e-05,
|
| 967 |
+
"loss": 1.7771,
|
| 968 |
+
"step": 680
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.3421578421578422,
|
| 972 |
+
"grad_norm": 0.15628665794752747,
|
| 973 |
+
"learning_rate": 1.664276886126246e-05,
|
| 974 |
+
"loss": 1.7738,
|
| 975 |
+
"step": 685
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.34465534465534464,
|
| 979 |
+
"grad_norm": 0.15746939136953078,
|
| 980 |
+
"learning_rate": 1.6577322739881255e-05,
|
| 981 |
+
"loss": 1.76,
|
| 982 |
+
"step": 690
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.34715284715284717,
|
| 986 |
+
"grad_norm": 0.1563776125400448,
|
| 987 |
+
"learning_rate": 1.6511376285984178e-05,
|
| 988 |
+
"loss": 1.7815,
|
| 989 |
+
"step": 695
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.34965034965034963,
|
| 993 |
+
"grad_norm": 0.15612599803803687,
|
| 994 |
+
"learning_rate": 1.6444934516073616e-05,
|
| 995 |
+
"loss": 1.7325,
|
| 996 |
+
"step": 700
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.35214785214785216,
|
| 1000 |
+
"grad_norm": 0.15539555825333878,
|
| 1001 |
+
"learning_rate": 1.6378002484330302e-05,
|
| 1002 |
+
"loss": 1.7267,
|
| 1003 |
+
"step": 705
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.3546453546453546,
|
| 1007 |
+
"grad_norm": 0.15742718718893856,
|
| 1008 |
+
"learning_rate": 1.631058528222886e-05,
|
| 1009 |
+
"loss": 1.7677,
|
| 1010 |
+
"step": 710
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.35714285714285715,
|
| 1014 |
+
"grad_norm": 0.15722376923699383,
|
| 1015 |
+
"learning_rate": 1.6242688038150506e-05,
|
| 1016 |
+
"loss": 1.7593,
|
| 1017 |
+
"step": 715
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.3596403596403596,
|
| 1021 |
+
"grad_norm": 0.15692354412727902,
|
| 1022 |
+
"learning_rate": 1.617431591699291e-05,
|
| 1023 |
+
"loss": 1.7681,
|
| 1024 |
+
"step": 720
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.36213786213786214,
|
| 1028 |
+
"grad_norm": 0.15962277561047927,
|
| 1029 |
+
"learning_rate": 1.610547411977734e-05,
|
| 1030 |
+
"loss": 1.75,
|
| 1031 |
+
"step": 725
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.3646353646353646,
|
| 1035 |
+
"grad_norm": 0.15692940824376528,
|
| 1036 |
+
"learning_rate": 1.6036167883252988e-05,
|
| 1037 |
+
"loss": 1.7444,
|
| 1038 |
+
"step": 730
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.36713286713286714,
|
| 1042 |
+
"grad_norm": 0.15649627153834378,
|
| 1043 |
+
"learning_rate": 1.5966402479498642e-05,
|
| 1044 |
+
"loss": 1.7191,
|
| 1045 |
+
"step": 735
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.3696303696303696,
|
| 1049 |
+
"grad_norm": 0.1573682637472189,
|
| 1050 |
+
"learning_rate": 1.589618321552163e-05,
|
| 1051 |
+
"loss": 1.78,
|
| 1052 |
+
"step": 740
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.37212787212787213,
|
| 1056 |
+
"grad_norm": 0.15762121053782066,
|
| 1057 |
+
"learning_rate": 1.5825515432854125e-05,
|
| 1058 |
+
"loss": 1.7795,
|
| 1059 |
+
"step": 745
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.37462537462537465,
|
| 1063 |
+
"grad_norm": 0.15776485071561994,
|
| 1064 |
+
"learning_rate": 1.575440450714681e-05,
|
| 1065 |
+
"loss": 1.7656,
|
| 1066 |
+
"step": 750
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.3771228771228771,
|
| 1070 |
+
"grad_norm": 0.15837437348496275,
|
| 1071 |
+
"learning_rate": 1.5682855847759965e-05,
|
| 1072 |
+
"loss": 1.7684,
|
| 1073 |
+
"step": 755
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.37962037962037964,
|
| 1077 |
+
"grad_norm": 0.16037425169330838,
|
| 1078 |
+
"learning_rate": 1.561087489735197e-05,
|
| 1079 |
+
"loss": 1.7368,
|
| 1080 |
+
"step": 760
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.3821178821178821,
|
| 1084 |
+
"grad_norm": 0.15513596949144706,
|
| 1085 |
+
"learning_rate": 1.55384671314653e-05,
|
| 1086 |
+
"loss": 1.7818,
|
| 1087 |
+
"step": 765
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.38461538461538464,
|
| 1091 |
+
"grad_norm": 0.15507828031304494,
|
| 1092 |
+
"learning_rate": 1.5465638058109998e-05,
|
| 1093 |
+
"loss": 1.7796,
|
| 1094 |
+
"step": 770
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.3871128871128871,
|
| 1098 |
+
"grad_norm": 0.1554815328421005,
|
| 1099 |
+
"learning_rate": 1.5392393217344666e-05,
|
| 1100 |
+
"loss": 1.755,
|
| 1101 |
+
"step": 775
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.38961038961038963,
|
| 1105 |
+
"grad_norm": 0.15688136225119317,
|
| 1106 |
+
"learning_rate": 1.5318738180855073e-05,
|
| 1107 |
+
"loss": 1.7387,
|
| 1108 |
+
"step": 780
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.3921078921078921,
|
| 1112 |
+
"grad_norm": 0.1582625776474904,
|
| 1113 |
+
"learning_rate": 1.5244678551530294e-05,
|
| 1114 |
+
"loss": 1.7603,
|
| 1115 |
+
"step": 785
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.3946053946053946,
|
| 1119 |
+
"grad_norm": 0.15701117307543766,
|
| 1120 |
+
"learning_rate": 1.5170219963036501e-05,
|
| 1121 |
+
"loss": 1.7182,
|
| 1122 |
+
"step": 790
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.3971028971028971,
|
| 1126 |
+
"grad_norm": 0.15766319963371506,
|
| 1127 |
+
"learning_rate": 1.5095368079388433e-05,
|
| 1128 |
+
"loss": 1.7295,
|
| 1129 |
+
"step": 795
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.3996003996003996,
|
| 1133 |
+
"grad_norm": 0.1587394631264011,
|
| 1134 |
+
"learning_rate": 1.5020128594518521e-05,
|
| 1135 |
+
"loss": 1.7914,
|
| 1136 |
+
"step": 800
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.4020979020979021,
|
| 1140 |
+
"grad_norm": 0.15654000030100182,
|
| 1141 |
+
"learning_rate": 1.4944507231843756e-05,
|
| 1142 |
+
"loss": 1.7202,
|
| 1143 |
+
"step": 805
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.4045954045954046,
|
| 1147 |
+
"grad_norm": 0.15292942125756898,
|
| 1148 |
+
"learning_rate": 1.4868509743830332e-05,
|
| 1149 |
+
"loss": 1.7901,
|
| 1150 |
+
"step": 810
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.4070929070929071,
|
| 1154 |
+
"grad_norm": 0.15560456619792792,
|
| 1155 |
+
"learning_rate": 1.4792141911556027e-05,
|
| 1156 |
+
"loss": 1.7716,
|
| 1157 |
+
"step": 815
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.4095904095904096,
|
| 1161 |
+
"grad_norm": 0.15543386661085493,
|
| 1162 |
+
"learning_rate": 1.4715409544270462e-05,
|
| 1163 |
+
"loss": 1.7203,
|
| 1164 |
+
"step": 820
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.41208791208791207,
|
| 1168 |
+
"grad_norm": 0.15809116648034297,
|
| 1169 |
+
"learning_rate": 1.4638318478953202e-05,
|
| 1170 |
+
"loss": 1.7382,
|
| 1171 |
+
"step": 825
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.4145854145854146,
|
| 1175 |
+
"grad_norm": 0.1557252652086651,
|
| 1176 |
+
"learning_rate": 1.4560874579869716e-05,
|
| 1177 |
+
"loss": 1.7729,
|
| 1178 |
+
"step": 830
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.41708291708291706,
|
| 1182 |
+
"grad_norm": 0.1558675093074507,
|
| 1183 |
+
"learning_rate": 1.4483083738125312e-05,
|
| 1184 |
+
"loss": 1.7453,
|
| 1185 |
+
"step": 835
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.4195804195804196,
|
| 1189 |
+
"grad_norm": 0.15507446312985207,
|
| 1190 |
+
"learning_rate": 1.440495187121698e-05,
|
| 1191 |
+
"loss": 1.7678,
|
| 1192 |
+
"step": 840
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.42207792207792205,
|
| 1196 |
+
"grad_norm": 0.1553337193710422,
|
| 1197 |
+
"learning_rate": 1.4326484922583277e-05,
|
| 1198 |
+
"loss": 1.7709,
|
| 1199 |
+
"step": 845
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.4245754245754246,
|
| 1203 |
+
"grad_norm": 0.15761636621288116,
|
| 1204 |
+
"learning_rate": 1.4247688861152195e-05,
|
| 1205 |
+
"loss": 1.7315,
|
| 1206 |
+
"step": 850
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.4270729270729271,
|
| 1210 |
+
"grad_norm": 0.15356881599814767,
|
| 1211 |
+
"learning_rate": 1.4168569680887115e-05,
|
| 1212 |
+
"loss": 1.6987,
|
| 1213 |
+
"step": 855
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.42957042957042957,
|
| 1217 |
+
"grad_norm": 0.15728107221945523,
|
| 1218 |
+
"learning_rate": 1.4089133400330851e-05,
|
| 1219 |
+
"loss": 1.741,
|
| 1220 |
+
"step": 860
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.4320679320679321,
|
| 1224 |
+
"grad_norm": 0.1548348784145128,
|
| 1225 |
+
"learning_rate": 1.4009386062147829e-05,
|
| 1226 |
+
"loss": 1.7323,
|
| 1227 |
+
"step": 865
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.43456543456543456,
|
| 1231 |
+
"grad_norm": 0.15572598457585934,
|
| 1232 |
+
"learning_rate": 1.3929333732664405e-05,
|
| 1233 |
+
"loss": 1.727,
|
| 1234 |
+
"step": 870
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.4370629370629371,
|
| 1238 |
+
"grad_norm": 0.15491152138136058,
|
| 1239 |
+
"learning_rate": 1.3848982501407433e-05,
|
| 1240 |
+
"loss": 1.7137,
|
| 1241 |
+
"step": 875
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.43956043956043955,
|
| 1245 |
+
"grad_norm": 0.15288889483520682,
|
| 1246 |
+
"learning_rate": 1.376833848064102e-05,
|
| 1247 |
+
"loss": 1.7241,
|
| 1248 |
+
"step": 880
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.4420579420579421,
|
| 1252 |
+
"grad_norm": 0.15423204666405854,
|
| 1253 |
+
"learning_rate": 1.3687407804901562e-05,
|
| 1254 |
+
"loss": 1.7536,
|
| 1255 |
+
"step": 885
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.44455544455544455,
|
| 1259 |
+
"grad_norm": 0.15495163216179897,
|
| 1260 |
+
"learning_rate": 1.3606196630531125e-05,
|
| 1261 |
+
"loss": 1.7184,
|
| 1262 |
+
"step": 890
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.44705294705294707,
|
| 1266 |
+
"grad_norm": 0.15761280747235365,
|
| 1267 |
+
"learning_rate": 1.3524711135209106e-05,
|
| 1268 |
+
"loss": 1.7751,
|
| 1269 |
+
"step": 895
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.44955044955044954,
|
| 1273 |
+
"grad_norm": 0.15841771000224297,
|
| 1274 |
+
"learning_rate": 1.344295751748231e-05,
|
| 1275 |
+
"loss": 1.7514,
|
| 1276 |
+
"step": 900
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.45204795204795206,
|
| 1280 |
+
"grad_norm": 0.15428205907298273,
|
| 1281 |
+
"learning_rate": 1.3360941996293439e-05,
|
| 1282 |
+
"loss": 1.716,
|
| 1283 |
+
"step": 905
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.45454545454545453,
|
| 1287 |
+
"grad_norm": 0.15303631590090105,
|
| 1288 |
+
"learning_rate": 1.3278670810508016e-05,
|
| 1289 |
+
"loss": 1.7083,
|
| 1290 |
+
"step": 910
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.45704295704295705,
|
| 1294 |
+
"grad_norm": 0.152891460359696,
|
| 1295 |
+
"learning_rate": 1.3196150218439791e-05,
|
| 1296 |
+
"loss": 1.711,
|
| 1297 |
+
"step": 915
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.4595404595404595,
|
| 1301 |
+
"grad_norm": 0.1541097644604054,
|
| 1302 |
+
"learning_rate": 1.3113386497374691e-05,
|
| 1303 |
+
"loss": 1.7352,
|
| 1304 |
+
"step": 920
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.46203796203796205,
|
| 1308 |
+
"grad_norm": 0.1581983359025658,
|
| 1309 |
+
"learning_rate": 1.303038594309329e-05,
|
| 1310 |
+
"loss": 1.7367,
|
| 1311 |
+
"step": 925
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.4645354645354645,
|
| 1315 |
+
"grad_norm": 0.15475095934873914,
|
| 1316 |
+
"learning_rate": 1.294715486939192e-05,
|
| 1317 |
+
"loss": 1.7169,
|
| 1318 |
+
"step": 930
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.46703296703296704,
|
| 1322 |
+
"grad_norm": 0.15844968557859115,
|
| 1323 |
+
"learning_rate": 1.2863699607602358e-05,
|
| 1324 |
+
"loss": 1.7208,
|
| 1325 |
+
"step": 935
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.4695304695304695,
|
| 1329 |
+
"grad_norm": 0.15639170445946177,
|
| 1330 |
+
"learning_rate": 1.2780026506110225e-05,
|
| 1331 |
+
"loss": 1.7242,
|
| 1332 |
+
"step": 940
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.47202797202797203,
|
| 1336 |
+
"grad_norm": 0.15698004903219,
|
| 1337 |
+
"learning_rate": 1.2696141929872064e-05,
|
| 1338 |
+
"loss": 1.754,
|
| 1339 |
+
"step": 945
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.4745254745254745,
|
| 1343 |
+
"grad_norm": 0.15510251460198132,
|
| 1344 |
+
"learning_rate": 1.2612052259931146e-05,
|
| 1345 |
+
"loss": 1.7437,
|
| 1346 |
+
"step": 950
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.477022977022977,
|
| 1350 |
+
"grad_norm": 0.1569119958304332,
|
| 1351 |
+
"learning_rate": 1.2527763892932098e-05,
|
| 1352 |
+
"loss": 1.7334,
|
| 1353 |
+
"step": 955
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.47952047952047955,
|
| 1357 |
+
"grad_norm": 0.15742453447805246,
|
| 1358 |
+
"learning_rate": 1.2443283240634295e-05,
|
| 1359 |
+
"loss": 1.7468,
|
| 1360 |
+
"step": 960
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.482017982017982,
|
| 1364 |
+
"grad_norm": 0.15752206416220757,
|
| 1365 |
+
"learning_rate": 1.2358616729424113e-05,
|
| 1366 |
+
"loss": 1.7285,
|
| 1367 |
+
"step": 965
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.48451548451548454,
|
| 1371 |
+
"grad_norm": 0.15404918669113854,
|
| 1372 |
+
"learning_rate": 1.2273770799826104e-05,
|
| 1373 |
+
"loss": 1.6984,
|
| 1374 |
+
"step": 970
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.487012987012987,
|
| 1378 |
+
"grad_norm": 0.15592357368342188,
|
| 1379 |
+
"learning_rate": 1.2188751906013054e-05,
|
| 1380 |
+
"loss": 1.7329,
|
| 1381 |
+
"step": 975
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.48951048951048953,
|
| 1385 |
+
"grad_norm": 0.156094008256334,
|
| 1386 |
+
"learning_rate": 1.2103566515315015e-05,
|
| 1387 |
+
"loss": 1.7513,
|
| 1388 |
+
"step": 980
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.492007992007992,
|
| 1392 |
+
"grad_norm": 0.1563990934117787,
|
| 1393 |
+
"learning_rate": 1.2018221107727348e-05,
|
| 1394 |
+
"loss": 1.7353,
|
| 1395 |
+
"step": 985
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.4945054945054945,
|
| 1399 |
+
"grad_norm": 0.1562707992028513,
|
| 1400 |
+
"learning_rate": 1.1932722175417796e-05,
|
| 1401 |
+
"loss": 1.7572,
|
| 1402 |
+
"step": 990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.497002997002997,
|
| 1406 |
+
"grad_norm": 0.1535470008052979,
|
| 1407 |
+
"learning_rate": 1.1847076222232614e-05,
|
| 1408 |
+
"loss": 1.7176,
|
| 1409 |
+
"step": 995
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.4995004995004995,
|
| 1413 |
+
"grad_norm": 0.1582614864687129,
|
| 1414 |
+
"learning_rate": 1.1761289763201843e-05,
|
| 1415 |
+
"loss": 1.7784,
|
| 1416 |
+
"step": 1000
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.501998001998002,
|
| 1420 |
+
"grad_norm": 0.155758931237908,
|
| 1421 |
+
"learning_rate": 1.1675369324043711e-05,
|
| 1422 |
+
"loss": 1.7144,
|
| 1423 |
+
"step": 1005
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.5044955044955045,
|
| 1427 |
+
"grad_norm": 0.15984802366682394,
|
| 1428 |
+
"learning_rate": 1.1589321440668215e-05,
|
| 1429 |
+
"loss": 1.7144,
|
| 1430 |
+
"step": 1010
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.506993006993007,
|
| 1434 |
+
"grad_norm": 0.1593119494177456,
|
| 1435 |
+
"learning_rate": 1.150315265867996e-05,
|
| 1436 |
+
"loss": 1.6995,
|
| 1437 |
+
"step": 1015
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.5094905094905094,
|
| 1441 |
+
"grad_norm": 0.15572494789848285,
|
| 1442 |
+
"learning_rate": 1.1416869532880219e-05,
|
| 1443 |
+
"loss": 1.7289,
|
| 1444 |
+
"step": 1020
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.511988011988012,
|
| 1448 |
+
"grad_norm": 0.15289377290260284,
|
| 1449 |
+
"learning_rate": 1.1330478626768322e-05,
|
| 1450 |
+
"loss": 1.7166,
|
| 1451 |
+
"step": 1025
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.5144855144855145,
|
| 1455 |
+
"grad_norm": 0.15398889595667828,
|
| 1456 |
+
"learning_rate": 1.1243986512042377e-05,
|
| 1457 |
+
"loss": 1.7163,
|
| 1458 |
+
"step": 1030
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.516983016983017,
|
| 1462 |
+
"grad_norm": 0.15490392246805634,
|
| 1463 |
+
"learning_rate": 1.1157399768099366e-05,
|
| 1464 |
+
"loss": 1.7134,
|
| 1465 |
+
"step": 1035
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.5194805194805194,
|
| 1469 |
+
"grad_norm": 0.15500679881154095,
|
| 1470 |
+
"learning_rate": 1.1070724981534647e-05,
|
| 1471 |
+
"loss": 1.7396,
|
| 1472 |
+
"step": 1040
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.521978021978022,
|
| 1476 |
+
"grad_norm": 0.1560235269544757,
|
| 1477 |
+
"learning_rate": 1.0983968745640923e-05,
|
| 1478 |
+
"loss": 1.6969,
|
| 1479 |
+
"step": 1045
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.5244755244755245,
|
| 1483 |
+
"grad_norm": 0.15453453742882894,
|
| 1484 |
+
"learning_rate": 1.0897137659906688e-05,
|
| 1485 |
+
"loss": 1.7426,
|
| 1486 |
+
"step": 1050
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.526973026973027,
|
| 1490 |
+
"grad_norm": 0.15599081676361193,
|
| 1491 |
+
"learning_rate": 1.0810238329514225e-05,
|
| 1492 |
+
"loss": 1.7457,
|
| 1493 |
+
"step": 1055
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.5294705294705294,
|
| 1497 |
+
"grad_norm": 0.15738103416797297,
|
| 1498 |
+
"learning_rate": 1.072327736483713e-05,
|
| 1499 |
+
"loss": 1.7159,
|
| 1500 |
+
"step": 1060
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.531968031968032,
|
| 1504 |
+
"grad_norm": 0.15854287836067726,
|
| 1505 |
+
"learning_rate": 1.0636261380937483e-05,
|
| 1506 |
+
"loss": 1.7684,
|
| 1507 |
+
"step": 1065
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.5344655344655345,
|
| 1511 |
+
"grad_norm": 0.15666913409437308,
|
| 1512 |
+
"learning_rate": 1.0549196997062636e-05,
|
| 1513 |
+
"loss": 1.7533,
|
| 1514 |
+
"step": 1070
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.5369630369630369,
|
| 1518 |
+
"grad_norm": 0.15513752052678667,
|
| 1519 |
+
"learning_rate": 1.0462090836141705e-05,
|
| 1520 |
+
"loss": 1.7485,
|
| 1521 |
+
"step": 1075
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.5394605394605395,
|
| 1525 |
+
"grad_norm": 0.15606650567355096,
|
| 1526 |
+
"learning_rate": 1.0374949524281741e-05,
|
| 1527 |
+
"loss": 1.7112,
|
| 1528 |
+
"step": 1080
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.541958041958042,
|
| 1532 |
+
"grad_norm": 0.15696280237552776,
|
| 1533 |
+
"learning_rate": 1.0287779690263718e-05,
|
| 1534 |
+
"loss": 1.7299,
|
| 1535 |
+
"step": 1085
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.5444555444555444,
|
| 1539 |
+
"grad_norm": 0.15466947191164923,
|
| 1540 |
+
"learning_rate": 1.020058796503826e-05,
|
| 1541 |
+
"loss": 1.7416,
|
| 1542 |
+
"step": 1090
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.5469530469530469,
|
| 1546 |
+
"grad_norm": 0.15659516724914946,
|
| 1547 |
+
"learning_rate": 1.0113380981221251e-05,
|
| 1548 |
+
"loss": 1.7309,
|
| 1549 |
+
"step": 1095
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.5494505494505495,
|
| 1553 |
+
"grad_norm": 0.15698307016673457,
|
| 1554 |
+
"learning_rate": 1.0026165372589278e-05,
|
| 1555 |
+
"loss": 1.7416,
|
| 1556 |
+
"step": 1100
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.551948051948052,
|
| 1560 |
+
"grad_norm": 0.15747483213576682,
|
| 1561 |
+
"learning_rate": 9.938947773575014e-06,
|
| 1562 |
+
"loss": 1.7354,
|
| 1563 |
+
"step": 1105
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.5544455544455544,
|
| 1567 |
+
"grad_norm": 0.1548555078024705,
|
| 1568 |
+
"learning_rate": 9.85173481876254e-06,
|
| 1569 |
+
"loss": 1.7285,
|
| 1570 |
+
"step": 1110
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.5569430569430569,
|
| 1574 |
+
"grad_norm": 0.1576329932512269,
|
| 1575 |
+
"learning_rate": 9.76453314238266e-06,
|
| 1576 |
+
"loss": 1.7295,
|
| 1577 |
+
"step": 1115
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.5594405594405595,
|
| 1581 |
+
"grad_norm": 0.15518997199188608,
|
| 1582 |
+
"learning_rate": 9.677349377808223e-06,
|
| 1583 |
+
"loss": 1.7252,
|
| 1584 |
+
"step": 1120
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.561938061938062,
|
| 1588 |
+
"grad_norm": 0.15573012190773938,
|
| 1589 |
+
"learning_rate": 9.59019015704955e-06,
|
| 1590 |
+
"loss": 1.7257,
|
| 1591 |
+
"step": 1125
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.5644355644355644,
|
| 1595 |
+
"grad_norm": 0.15100601160295765,
|
| 1596 |
+
"learning_rate": 9.50306211024993e-06,
|
| 1597 |
+
"loss": 1.6836,
|
| 1598 |
+
"step": 1130
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.5669330669330669,
|
| 1602 |
+
"grad_norm": 0.15501908738813838,
|
| 1603 |
+
"learning_rate": 9.415971865181262e-06,
|
| 1604 |
+
"loss": 1.7429,
|
| 1605 |
+
"step": 1135
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.5694305694305695,
|
| 1609 |
+
"grad_norm": 0.15812904061992633,
|
| 1610 |
+
"learning_rate": 9.328926046739899e-06,
|
| 1611 |
+
"loss": 1.7084,
|
| 1612 |
+
"step": 1140
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.5719280719280719,
|
| 1616 |
+
"grad_norm": 0.15627757158949415,
|
| 1617 |
+
"learning_rate": 9.241931276442692e-06,
|
| 1618 |
+
"loss": 1.7214,
|
| 1619 |
+
"step": 1145
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.5744255744255744,
|
| 1623 |
+
"grad_norm": 0.1558930669628705,
|
| 1624 |
+
"learning_rate": 9.154994171923285e-06,
|
| 1625 |
+
"loss": 1.7109,
|
| 1626 |
+
"step": 1150
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.5769230769230769,
|
| 1630 |
+
"grad_norm": 0.15441900101502826,
|
| 1631 |
+
"learning_rate": 9.068121346428735e-06,
|
| 1632 |
+
"loss": 1.7376,
|
| 1633 |
+
"step": 1155
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.5794205794205795,
|
| 1637 |
+
"grad_norm": 0.15623145805295774,
|
| 1638 |
+
"learning_rate": 8.981319408316435e-06,
|
| 1639 |
+
"loss": 1.736,
|
| 1640 |
+
"step": 1160
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.5819180819180819,
|
| 1644 |
+
"grad_norm": 0.15421994986436063,
|
| 1645 |
+
"learning_rate": 8.894594960551417e-06,
|
| 1646 |
+
"loss": 1.7484,
|
| 1647 |
+
"step": 1165
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.5844155844155844,
|
| 1651 |
+
"grad_norm": 0.15562604285522325,
|
| 1652 |
+
"learning_rate": 8.807954600204079e-06,
|
| 1653 |
+
"loss": 1.7608,
|
| 1654 |
+
"step": 1170
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.586913086913087,
|
| 1658 |
+
"grad_norm": 0.15592693727619433,
|
| 1659 |
+
"learning_rate": 8.721404917948343e-06,
|
| 1660 |
+
"loss": 1.7358,
|
| 1661 |
+
"step": 1175
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.5894105894105894,
|
| 1665 |
+
"grad_norm": 0.15213420323940668,
|
| 1666 |
+
"learning_rate": 8.634952497560311e-06,
|
| 1667 |
+
"loss": 1.6824,
|
| 1668 |
+
"step": 1180
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.5919080919080919,
|
| 1672 |
+
"grad_norm": 0.15353479909518586,
|
| 1673 |
+
"learning_rate": 8.548603915417445e-06,
|
| 1674 |
+
"loss": 1.7362,
|
| 1675 |
+
"step": 1185
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.5944055944055944,
|
| 1679 |
+
"grad_norm": 0.1568536405873683,
|
| 1680 |
+
"learning_rate": 8.462365739998293e-06,
|
| 1681 |
+
"loss": 1.7144,
|
| 1682 |
+
"step": 1190
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.596903096903097,
|
| 1686 |
+
"grad_norm": 0.15792171997932242,
|
| 1687 |
+
"learning_rate": 8.37624453138284e-06,
|
| 1688 |
+
"loss": 1.7364,
|
| 1689 |
+
"step": 1195
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.5994005994005994,
|
| 1693 |
+
"grad_norm": 0.1559386534693386,
|
| 1694 |
+
"learning_rate": 8.29024684075349e-06,
|
| 1695 |
+
"loss": 1.7593,
|
| 1696 |
+
"step": 1200
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.6018981018981019,
|
| 1700 |
+
"grad_norm": 0.15715611634818757,
|
| 1701 |
+
"learning_rate": 8.204379209896712e-06,
|
| 1702 |
+
"loss": 1.7032,
|
| 1703 |
+
"step": 1205
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.6043956043956044,
|
| 1707 |
+
"grad_norm": 0.15372344898616952,
|
| 1708 |
+
"learning_rate": 8.118648170705418e-06,
|
| 1709 |
+
"loss": 1.6978,
|
| 1710 |
+
"step": 1210
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.6068931068931069,
|
| 1714 |
+
"grad_norm": 0.15810145378733176,
|
| 1715 |
+
"learning_rate": 8.033060244682079e-06,
|
| 1716 |
+
"loss": 1.7396,
|
| 1717 |
+
"step": 1215
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.6093906093906094,
|
| 1721 |
+
"grad_norm": 0.1557299319735277,
|
| 1722 |
+
"learning_rate": 7.947621942442651e-06,
|
| 1723 |
+
"loss": 1.7232,
|
| 1724 |
+
"step": 1220
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.6118881118881119,
|
| 1728 |
+
"grad_norm": 0.1557147874512252,
|
| 1729 |
+
"learning_rate": 7.86233976322131e-06,
|
| 1730 |
+
"loss": 1.71,
|
| 1731 |
+
"step": 1225
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.6143856143856143,
|
| 1735 |
+
"grad_norm": 0.1554308477187568,
|
| 1736 |
+
"learning_rate": 7.777220194376047e-06,
|
| 1737 |
+
"loss": 1.7177,
|
| 1738 |
+
"step": 1230
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.6168831168831169,
|
| 1742 |
+
"grad_norm": 0.1564865748636126,
|
| 1743 |
+
"learning_rate": 7.692269710895204e-06,
|
| 1744 |
+
"loss": 1.705,
|
| 1745 |
+
"step": 1235
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.6193806193806194,
|
| 1749 |
+
"grad_norm": 0.1540720670994077,
|
| 1750 |
+
"learning_rate": 7.6074947749049085e-06,
|
| 1751 |
+
"loss": 1.7039,
|
| 1752 |
+
"step": 1240
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.6218781218781219,
|
| 1756 |
+
"grad_norm": 0.15756506692256836,
|
| 1757 |
+
"learning_rate": 7.522901835177506e-06,
|
| 1758 |
+
"loss": 1.7217,
|
| 1759 |
+
"step": 1245
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.6243756243756243,
|
| 1763 |
+
"grad_norm": 0.15371879475088404,
|
| 1764 |
+
"learning_rate": 7.438497326641012e-06,
|
| 1765 |
+
"loss": 1.6895,
|
| 1766 |
+
"step": 1250
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.6268731268731269,
|
| 1770 |
+
"grad_norm": 0.15945775556289293,
|
| 1771 |
+
"learning_rate": 7.354287669889606e-06,
|
| 1772 |
+
"loss": 1.7366,
|
| 1773 |
+
"step": 1255
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.6293706293706294,
|
| 1777 |
+
"grad_norm": 0.15778707821642424,
|
| 1778 |
+
"learning_rate": 7.270279270695224e-06,
|
| 1779 |
+
"loss": 1.7303,
|
| 1780 |
+
"step": 1260
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.6318681318681318,
|
| 1784 |
+
"grad_norm": 0.1586505017667703,
|
| 1785 |
+
"learning_rate": 7.186478519520279e-06,
|
| 1786 |
+
"loss": 1.7147,
|
| 1787 |
+
"step": 1265
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.6343656343656343,
|
| 1791 |
+
"grad_norm": 0.15789741519311937,
|
| 1792 |
+
"learning_rate": 7.10289179103153e-06,
|
| 1793 |
+
"loss": 1.7146,
|
| 1794 |
+
"step": 1270
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.6368631368631369,
|
| 1798 |
+
"grad_norm": 0.15184374761336888,
|
| 1799 |
+
"learning_rate": 7.019525443615181e-06,
|
| 1800 |
+
"loss": 1.7195,
|
| 1801 |
+
"step": 1275
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.6393606393606394,
|
| 1805 |
+
"grad_norm": 0.15637777552677476,
|
| 1806 |
+
"learning_rate": 6.936385818893197e-06,
|
| 1807 |
+
"loss": 1.7471,
|
| 1808 |
+
"step": 1280
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.6418581418581418,
|
| 1812 |
+
"grad_norm": 0.1567686671030996,
|
| 1813 |
+
"learning_rate": 6.853479241240895e-06,
|
| 1814 |
+
"loss": 1.7436,
|
| 1815 |
+
"step": 1285
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.6443556443556444,
|
| 1819 |
+
"grad_norm": 0.15591409064699058,
|
| 1820 |
+
"learning_rate": 6.770812017305862e-06,
|
| 1821 |
+
"loss": 1.7252,
|
| 1822 |
+
"step": 1290
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.6468531468531469,
|
| 1826 |
+
"grad_norm": 0.15804342302479574,
|
| 1827 |
+
"learning_rate": 6.688390435528209e-06,
|
| 1828 |
+
"loss": 1.7334,
|
| 1829 |
+
"step": 1295
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.6493506493506493,
|
| 1833 |
+
"grad_norm": 0.15436002266775298,
|
| 1834 |
+
"learning_rate": 6.606220765662209e-06,
|
| 1835 |
+
"loss": 1.7165,
|
| 1836 |
+
"step": 1300
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.6518481518481518,
|
| 1840 |
+
"grad_norm": 0.15531296312132536,
|
| 1841 |
+
"learning_rate": 6.524309258299368e-06,
|
| 1842 |
+
"loss": 1.7605,
|
| 1843 |
+
"step": 1305
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.6543456543456544,
|
| 1847 |
+
"grad_norm": 0.1565211383258821,
|
| 1848 |
+
"learning_rate": 6.4426621443929505e-06,
|
| 1849 |
+
"loss": 1.7061,
|
| 1850 |
+
"step": 1310
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.6568431568431569,
|
| 1854 |
+
"grad_norm": 0.1582362713382269,
|
| 1855 |
+
"learning_rate": 6.361285634783976e-06,
|
| 1856 |
+
"loss": 1.7196,
|
| 1857 |
+
"step": 1315
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.6593406593406593,
|
| 1861 |
+
"grad_norm": 0.15521707003590374,
|
| 1862 |
+
"learning_rate": 6.280185919728784e-06,
|
| 1863 |
+
"loss": 1.6963,
|
| 1864 |
+
"step": 1320
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.6618381618381618,
|
| 1868 |
+
"grad_norm": 0.15705101318113973,
|
| 1869 |
+
"learning_rate": 6.199369168428143e-06,
|
| 1870 |
+
"loss": 1.742,
|
| 1871 |
+
"step": 1325
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.6643356643356644,
|
| 1875 |
+
"grad_norm": 0.15356334771300495,
|
| 1876 |
+
"learning_rate": 6.11884152855795e-06,
|
| 1877 |
+
"loss": 1.7011,
|
| 1878 |
+
"step": 1330
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.6668331668331668,
|
| 1882 |
+
"grad_norm": 0.15700312151436902,
|
| 1883 |
+
"learning_rate": 6.0386091258015965e-06,
|
| 1884 |
+
"loss": 1.6928,
|
| 1885 |
+
"step": 1335
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.6693306693306693,
|
| 1889 |
+
"grad_norm": 0.15784510931828966,
|
| 1890 |
+
"learning_rate": 5.95867806338398e-06,
|
| 1891 |
+
"loss": 1.7009,
|
| 1892 |
+
"step": 1340
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.6718281718281718,
|
| 1896 |
+
"grad_norm": 0.15746608991343575,
|
| 1897 |
+
"learning_rate": 5.879054421607248e-06,
|
| 1898 |
+
"loss": 1.7397,
|
| 1899 |
+
"step": 1345
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.6743256743256744,
|
| 1903 |
+
"grad_norm": 0.15620423042692408,
|
| 1904 |
+
"learning_rate": 5.79974425738826e-06,
|
| 1905 |
+
"loss": 1.7045,
|
| 1906 |
+
"step": 1350
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.6768231768231768,
|
| 1910 |
+
"grad_norm": 0.15370013592929113,
|
| 1911 |
+
"learning_rate": 5.720753603797855e-06,
|
| 1912 |
+
"loss": 1.7015,
|
| 1913 |
+
"step": 1355
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.6793206793206793,
|
| 1917 |
+
"grad_norm": 0.15590449648570684,
|
| 1918 |
+
"learning_rate": 5.6420884696019085e-06,
|
| 1919 |
+
"loss": 1.6937,
|
| 1920 |
+
"step": 1360
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.6818181818181818,
|
| 1924 |
+
"grad_norm": 0.15510936819450175,
|
| 1925 |
+
"learning_rate": 5.563754838804252e-06,
|
| 1926 |
+
"loss": 1.7019,
|
| 1927 |
+
"step": 1365
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.6843156843156843,
|
| 1931 |
+
"grad_norm": 0.1527676787399584,
|
| 1932 |
+
"learning_rate": 5.485758670191486e-06,
|
| 1933 |
+
"loss": 1.7051,
|
| 1934 |
+
"step": 1370
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.6868131868131868,
|
| 1938 |
+
"grad_norm": 0.15546653572707683,
|
| 1939 |
+
"learning_rate": 5.408105896879684e-06,
|
| 1940 |
+
"loss": 1.7296,
|
| 1941 |
+
"step": 1375
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.6893106893106893,
|
| 1945 |
+
"grad_norm": 0.15309923322143176,
|
| 1946 |
+
"learning_rate": 5.330802425863064e-06,
|
| 1947 |
+
"loss": 1.6989,
|
| 1948 |
+
"step": 1380
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.6918081918081919,
|
| 1952 |
+
"grad_norm": 0.15308524124271006,
|
| 1953 |
+
"learning_rate": 5.2538541375646675e-06,
|
| 1954 |
+
"loss": 1.6826,
|
| 1955 |
+
"step": 1385
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.6943056943056943,
|
| 1959 |
+
"grad_norm": 0.1538371045270346,
|
| 1960 |
+
"learning_rate": 5.177266885389016e-06,
|
| 1961 |
+
"loss": 1.6983,
|
| 1962 |
+
"step": 1390
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.6968031968031968,
|
| 1966 |
+
"grad_norm": 0.15413826922713267,
|
| 1967 |
+
"learning_rate": 5.101046495276852e-06,
|
| 1968 |
+
"loss": 1.724,
|
| 1969 |
+
"step": 1395
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.6993006993006993,
|
| 1973 |
+
"grad_norm": 0.15297221612329867,
|
| 1974 |
+
"learning_rate": 5.025198765261981e-06,
|
| 1975 |
+
"loss": 1.7201,
|
| 1976 |
+
"step": 1400
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.7017982017982018,
|
| 1980 |
+
"grad_norm": 0.15512908248946458,
|
| 1981 |
+
"learning_rate": 4.949729465030193e-06,
|
| 1982 |
+
"loss": 1.6795,
|
| 1983 |
+
"step": 1405
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.7042957042957043,
|
| 1987 |
+
"grad_norm": 0.15505906851804813,
|
| 1988 |
+
"learning_rate": 4.874644335480383e-06,
|
| 1989 |
+
"loss": 1.707,
|
| 1990 |
+
"step": 1410
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.7067932067932068,
|
| 1994 |
+
"grad_norm": 0.15397133300096347,
|
| 1995 |
+
"learning_rate": 4.799949088287849e-06,
|
| 1996 |
+
"loss": 1.7008,
|
| 1997 |
+
"step": 1415
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.7092907092907093,
|
| 2001 |
+
"grad_norm": 0.15813016475329747,
|
| 2002 |
+
"learning_rate": 4.725649405469786e-06,
|
| 2003 |
+
"loss": 1.7325,
|
| 2004 |
+
"step": 1420
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.7117882117882118,
|
| 2008 |
+
"grad_norm": 0.15446637709842673,
|
| 2009 |
+
"learning_rate": 4.65175093895308e-06,
|
| 2010 |
+
"loss": 1.7215,
|
| 2011 |
+
"step": 1425
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.7142857142857143,
|
| 2015 |
+
"grad_norm": 0.1561588030182068,
|
| 2016 |
+
"learning_rate": 4.57825931014437e-06,
|
| 2017 |
+
"loss": 1.7184,
|
| 2018 |
+
"step": 1430
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.7167832167832168,
|
| 2022 |
+
"grad_norm": 0.15772322144743187,
|
| 2023 |
+
"learning_rate": 4.50518010950241e-06,
|
| 2024 |
+
"loss": 1.7166,
|
| 2025 |
+
"step": 1435
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.7192807192807192,
|
| 2029 |
+
"grad_norm": 0.15727583874213139,
|
| 2030 |
+
"learning_rate": 4.432518896112831e-06,
|
| 2031 |
+
"loss": 1.7034,
|
| 2032 |
+
"step": 1440
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.7217782217782218,
|
| 2036 |
+
"grad_norm": 0.153411598788086,
|
| 2037 |
+
"learning_rate": 4.360281197265249e-06,
|
| 2038 |
+
"loss": 1.695,
|
| 2039 |
+
"step": 1445
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.7242757242757243,
|
| 2043 |
+
"grad_norm": 0.15655819818705863,
|
| 2044 |
+
"learning_rate": 4.2884725080328245e-06,
|
| 2045 |
+
"loss": 1.7411,
|
| 2046 |
+
"step": 1450
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.7267732267732268,
|
| 2050 |
+
"grad_norm": 0.15649221553246048,
|
| 2051 |
+
"learning_rate": 4.217098290854234e-06,
|
| 2052 |
+
"loss": 1.6852,
|
| 2053 |
+
"step": 1455
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.7292707292707292,
|
| 2057 |
+
"grad_norm": 0.15608752112425878,
|
| 2058 |
+
"learning_rate": 4.146163975118154e-06,
|
| 2059 |
+
"loss": 1.7158,
|
| 2060 |
+
"step": 1460
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.7317682317682318,
|
| 2064 |
+
"grad_norm": 0.156378012901712,
|
| 2065 |
+
"learning_rate": 4.0756749567502704e-06,
|
| 2066 |
+
"loss": 1.7141,
|
| 2067 |
+
"step": 1465
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.7342657342657343,
|
| 2071 |
+
"grad_norm": 0.15573008458523765,
|
| 2072 |
+
"learning_rate": 4.005636597802785e-06,
|
| 2073 |
+
"loss": 1.7316,
|
| 2074 |
+
"step": 1470
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.7367632367632367,
|
| 2078 |
+
"grad_norm": 0.15624579822166534,
|
| 2079 |
+
"learning_rate": 3.936054226046539e-06,
|
| 2080 |
+
"loss": 1.7275,
|
| 2081 |
+
"step": 1475
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.7392607392607392,
|
| 2085 |
+
"grad_norm": 0.15708652587903585,
|
| 2086 |
+
"learning_rate": 3.866933134565747e-06,
|
| 2087 |
+
"loss": 1.7146,
|
| 2088 |
+
"step": 1480
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.7417582417582418,
|
| 2092 |
+
"grad_norm": 0.15685542138951347,
|
| 2093 |
+
"learning_rate": 3.7982785813553335e-06,
|
| 2094 |
+
"loss": 1.7009,
|
| 2095 |
+
"step": 1485
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.7442557442557443,
|
| 2099 |
+
"grad_norm": 0.15672884263951484,
|
| 2100 |
+
"learning_rate": 3.730095788920969e-06,
|
| 2101 |
+
"loss": 1.7273,
|
| 2102 |
+
"step": 1490
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.7467532467532467,
|
| 2106 |
+
"grad_norm": 0.15717150554726544,
|
| 2107 |
+
"learning_rate": 3.662389943881811e-06,
|
| 2108 |
+
"loss": 1.7223,
|
| 2109 |
+
"step": 1495
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.7492507492507493,
|
| 2113 |
+
"grad_norm": 0.153594301470774,
|
| 2114 |
+
"learning_rate": 3.5951661965759376e-06,
|
| 2115 |
+
"loss": 1.7232,
|
| 2116 |
+
"step": 1500
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.7517482517482518,
|
| 2120 |
+
"grad_norm": 0.15489817985281926,
|
| 2121 |
+
"learning_rate": 3.5284296606685788e-06,
|
| 2122 |
+
"loss": 1.7198,
|
| 2123 |
+
"step": 1505
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.7542457542457542,
|
| 2127 |
+
"grad_norm": 0.15650041684511018,
|
| 2128 |
+
"learning_rate": 3.4621854127631293e-06,
|
| 2129 |
+
"loss": 1.7427,
|
| 2130 |
+
"step": 1510
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.7567432567432567,
|
| 2134 |
+
"grad_norm": 0.15414035664901124,
|
| 2135 |
+
"learning_rate": 3.3964384920149574e-06,
|
| 2136 |
+
"loss": 1.6846,
|
| 2137 |
+
"step": 1515
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.7592407592407593,
|
| 2141 |
+
"grad_norm": 0.15583094131363653,
|
| 2142 |
+
"learning_rate": 3.331193899748091e-06,
|
| 2143 |
+
"loss": 1.726,
|
| 2144 |
+
"step": 1520
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.7617382617382618,
|
| 2148 |
+
"grad_norm": 0.1554155623994627,
|
| 2149 |
+
"learning_rate": 3.2664565990747733e-06,
|
| 2150 |
+
"loss": 1.7206,
|
| 2151 |
+
"step": 1525
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.7642357642357642,
|
| 2155 |
+
"grad_norm": 0.15305845088804462,
|
| 2156 |
+
"learning_rate": 3.202231514517913e-06,
|
| 2157 |
+
"loss": 1.7028,
|
| 2158 |
+
"step": 1530
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.7667332667332667,
|
| 2162 |
+
"grad_norm": 0.15291334769441534,
|
| 2163 |
+
"learning_rate": 3.1385235316364805e-06,
|
| 2164 |
+
"loss": 1.7322,
|
| 2165 |
+
"step": 1535
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.7692307692307693,
|
| 2169 |
+
"grad_norm": 0.15661260187424764,
|
| 2170 |
+
"learning_rate": 3.0753374966538807e-06,
|
| 2171 |
+
"loss": 1.6848,
|
| 2172 |
+
"step": 1540
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.7717282717282717,
|
| 2176 |
+
"grad_norm": 0.15809358800271614,
|
| 2177 |
+
"learning_rate": 3.012678216089281e-06,
|
| 2178 |
+
"loss": 1.6997,
|
| 2179 |
+
"step": 1545
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.7742257742257742,
|
| 2183 |
+
"grad_norm": 0.155978034855165,
|
| 2184 |
+
"learning_rate": 2.9505504563920005e-06,
|
| 2185 |
+
"loss": 1.7356,
|
| 2186 |
+
"step": 1550
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.7767232767232767,
|
| 2190 |
+
"grad_norm": 0.15604151355576693,
|
| 2191 |
+
"learning_rate": 2.888958943578919e-06,
|
| 2192 |
+
"loss": 1.7165,
|
| 2193 |
+
"step": 1555
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.7792207792207793,
|
| 2197 |
+
"grad_norm": 0.1550171550409355,
|
| 2198 |
+
"learning_rate": 2.827908362874986e-06,
|
| 2199 |
+
"loss": 1.7025,
|
| 2200 |
+
"step": 1560
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.7817182817182817,
|
| 2204 |
+
"grad_norm": 0.15527890651905346,
|
| 2205 |
+
"learning_rate": 2.7674033583568004e-06,
|
| 2206 |
+
"loss": 1.687,
|
| 2207 |
+
"step": 1565
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.7842157842157842,
|
| 2211 |
+
"grad_norm": 0.15537639424177968,
|
| 2212 |
+
"learning_rate": 2.7074485325993482e-06,
|
| 2213 |
+
"loss": 1.7126,
|
| 2214 |
+
"step": 1570
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.7867132867132867,
|
| 2218 |
+
"grad_norm": 0.15244961389707662,
|
| 2219 |
+
"learning_rate": 2.648048446325894e-06,
|
| 2220 |
+
"loss": 1.6943,
|
| 2221 |
+
"step": 1575
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.7892107892107892,
|
| 2225 |
+
"grad_norm": 0.15407223657879413,
|
| 2226 |
+
"learning_rate": 2.5892076180610372e-06,
|
| 2227 |
+
"loss": 1.7077,
|
| 2228 |
+
"step": 1580
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.7917082917082917,
|
| 2232 |
+
"grad_norm": 0.15395383095089893,
|
| 2233 |
+
"learning_rate": 2.5309305237869953e-06,
|
| 2234 |
+
"loss": 1.7428,
|
| 2235 |
+
"step": 1585
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.7942057942057942,
|
| 2239 |
+
"grad_norm": 0.15693221036515126,
|
| 2240 |
+
"learning_rate": 2.473221596603127e-06,
|
| 2241 |
+
"loss": 1.7003,
|
| 2242 |
+
"step": 1590
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.7967032967032966,
|
| 2246 |
+
"grad_norm": 0.15202864145333894,
|
| 2247 |
+
"learning_rate": 2.416085226388699e-06,
|
| 2248 |
+
"loss": 1.7062,
|
| 2249 |
+
"step": 1595
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.7992007992007992,
|
| 2253 |
+
"grad_norm": 0.15438777267790824,
|
| 2254 |
+
"learning_rate": 2.3595257594689504e-06,
|
| 2255 |
+
"loss": 1.716,
|
| 2256 |
+
"step": 1600
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.8016983016983017,
|
| 2260 |
+
"grad_norm": 0.15408101244601666,
|
| 2261 |
+
"learning_rate": 2.303547498284483e-06,
|
| 2262 |
+
"loss": 1.74,
|
| 2263 |
+
"step": 1605
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.8041958041958042,
|
| 2267 |
+
"grad_norm": 0.1529673037596131,
|
| 2268 |
+
"learning_rate": 2.2481547010639648e-06,
|
| 2269 |
+
"loss": 1.724,
|
| 2270 |
+
"step": 1610
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.8066933066933067,
|
| 2274 |
+
"grad_norm": 0.15429371566399036,
|
| 2275 |
+
"learning_rate": 2.1933515815002115e-06,
|
| 2276 |
+
"loss": 1.7058,
|
| 2277 |
+
"step": 1615
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.8091908091908092,
|
| 2281 |
+
"grad_norm": 0.1553802340727305,
|
| 2282 |
+
"learning_rate": 2.1391423084296627e-06,
|
| 2283 |
+
"loss": 1.7044,
|
| 2284 |
+
"step": 1620
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.8116883116883117,
|
| 2288 |
+
"grad_norm": 0.15552250374029558,
|
| 2289 |
+
"learning_rate": 2.08553100551525e-06,
|
| 2290 |
+
"loss": 1.7294,
|
| 2291 |
+
"step": 1625
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.8141858141858141,
|
| 2295 |
+
"grad_norm": 0.15556388073981073,
|
| 2296 |
+
"learning_rate": 2.0325217509327145e-06,
|
| 2297 |
+
"loss": 1.6825,
|
| 2298 |
+
"step": 1630
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.8166833166833167,
|
| 2302 |
+
"grad_norm": 0.1534237780794405,
|
| 2303 |
+
"learning_rate": 1.980118577060397e-06,
|
| 2304 |
+
"loss": 1.7258,
|
| 2305 |
+
"step": 1635
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.8191808191808192,
|
| 2309 |
+
"grad_norm": 0.1556601065527979,
|
| 2310 |
+
"learning_rate": 1.9283254701724742e-06,
|
| 2311 |
+
"loss": 1.6938,
|
| 2312 |
+
"step": 1640
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.8216783216783217,
|
| 2316 |
+
"grad_norm": 0.156089780593886,
|
| 2317 |
+
"learning_rate": 1.8771463701357428e-06,
|
| 2318 |
+
"loss": 1.7357,
|
| 2319 |
+
"step": 1645
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.8241758241758241,
|
| 2323 |
+
"grad_norm": 0.15554888194693464,
|
| 2324 |
+
"learning_rate": 1.8265851701099146e-06,
|
| 2325 |
+
"loss": 1.7063,
|
| 2326 |
+
"step": 1650
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.8266733266733267,
|
| 2330 |
+
"grad_norm": 0.15466801254433934,
|
| 2331 |
+
"learning_rate": 1.7766457162514594e-06,
|
| 2332 |
+
"loss": 1.7435,
|
| 2333 |
+
"step": 1655
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.8291708291708292,
|
| 2337 |
+
"grad_norm": 0.15476788949201128,
|
| 2338 |
+
"learning_rate": 1.7273318074210298e-06,
|
| 2339 |
+
"loss": 1.7059,
|
| 2340 |
+
"step": 1660
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.8316683316683317,
|
| 2344 |
+
"grad_norm": 0.15219152100759303,
|
| 2345 |
+
"learning_rate": 1.6786471948944994e-06,
|
| 2346 |
+
"loss": 1.6964,
|
| 2347 |
+
"step": 1665
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.8341658341658341,
|
| 2351 |
+
"grad_norm": 0.1548313560211368,
|
| 2352 |
+
"learning_rate": 1.630595582077591e-06,
|
| 2353 |
+
"loss": 1.7,
|
| 2354 |
+
"step": 1670
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.8366633366633367,
|
| 2358 |
+
"grad_norm": 0.15585812651840628,
|
| 2359 |
+
"learning_rate": 1.5831806242241632e-06,
|
| 2360 |
+
"loss": 1.7028,
|
| 2361 |
+
"step": 1675
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.8391608391608392,
|
| 2365 |
+
"grad_norm": 0.15443911839682758,
|
| 2366 |
+
"learning_rate": 1.5364059281581566e-06,
|
| 2367 |
+
"loss": 1.7176,
|
| 2368 |
+
"step": 1680
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.8416583416583416,
|
| 2372 |
+
"grad_norm": 0.1605017036197551,
|
| 2373 |
+
"learning_rate": 1.4902750519992392e-06,
|
| 2374 |
+
"loss": 1.7156,
|
| 2375 |
+
"step": 1685
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.8441558441558441,
|
| 2379 |
+
"grad_norm": 0.1538258207387918,
|
| 2380 |
+
"learning_rate": 1.4447915048921224e-06,
|
| 2381 |
+
"loss": 1.6744,
|
| 2382 |
+
"step": 1690
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.8466533466533467,
|
| 2386 |
+
"grad_norm": 0.15406565385641532,
|
| 2387 |
+
"learning_rate": 1.3999587467396336e-06,
|
| 2388 |
+
"loss": 1.6946,
|
| 2389 |
+
"step": 1695
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.8491508491508492,
|
| 2393 |
+
"grad_norm": 0.1536094454245134,
|
| 2394 |
+
"learning_rate": 1.3557801879395283e-06,
|
| 2395 |
+
"loss": 1.7148,
|
| 2396 |
+
"step": 1700
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.8516483516483516,
|
| 2400 |
+
"grad_norm": 0.15544099947612952,
|
| 2401 |
+
"learning_rate": 1.3122591891250492e-06,
|
| 2402 |
+
"loss": 1.6875,
|
| 2403 |
+
"step": 1705
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.8541458541458542,
|
| 2407 |
+
"grad_norm": 0.16188709049635025,
|
| 2408 |
+
"learning_rate": 1.2693990609092965e-06,
|
| 2409 |
+
"loss": 1.7387,
|
| 2410 |
+
"step": 1710
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.8566433566433567,
|
| 2414 |
+
"grad_norm": 0.15363848054893653,
|
| 2415 |
+
"learning_rate": 1.227203063633393e-06,
|
| 2416 |
+
"loss": 1.7277,
|
| 2417 |
+
"step": 1715
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.8591408591408591,
|
| 2421 |
+
"grad_norm": 0.15554467350943088,
|
| 2422 |
+
"learning_rate": 1.185674407118461e-06,
|
| 2423 |
+
"loss": 1.6892,
|
| 2424 |
+
"step": 1720
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.8616383616383616,
|
| 2428 |
+
"grad_norm": 0.15618845824953714,
|
| 2429 |
+
"learning_rate": 1.1448162504214621e-06,
|
| 2430 |
+
"loss": 1.7413,
|
| 2431 |
+
"step": 1725
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.8641358641358642,
|
| 2435 |
+
"grad_norm": 0.15459131494726566,
|
| 2436 |
+
"learning_rate": 1.1046317015948904e-06,
|
| 2437 |
+
"loss": 1.7133,
|
| 2438 |
+
"step": 1730
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.8666333666333667,
|
| 2442 |
+
"grad_norm": 0.1552140715983614,
|
| 2443 |
+
"learning_rate": 1.0651238174503408e-06,
|
| 2444 |
+
"loss": 1.7147,
|
| 2445 |
+
"step": 1735
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 0.8691308691308691,
|
| 2449 |
+
"grad_norm": 0.15520848465047085,
|
| 2450 |
+
"learning_rate": 1.0262956033259775e-06,
|
| 2451 |
+
"loss": 1.7367,
|
| 2452 |
+
"step": 1740
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.8716283716283716,
|
| 2456 |
+
"grad_norm": 0.15402498238996434,
|
| 2457 |
+
"learning_rate": 9.881500128579314e-07,
|
| 2458 |
+
"loss": 1.6949,
|
| 2459 |
+
"step": 1745
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.8741258741258742,
|
| 2463 |
+
"grad_norm": 0.15820953048710004,
|
| 2464 |
+
"learning_rate": 9.506899477556042e-07,
|
| 2465 |
+
"loss": 1.6936,
|
| 2466 |
+
"step": 1750
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 0.8766233766233766,
|
| 2470 |
+
"grad_norm": 0.15475714836078286,
|
| 2471 |
+
"learning_rate": 9.139182575809446e-07,
|
| 2472 |
+
"loss": 1.7216,
|
| 2473 |
+
"step": 1755
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.8791208791208791,
|
| 2477 |
+
"grad_norm": 0.15320851876078712,
|
| 2478 |
+
"learning_rate": 8.778377395316906e-07,
|
| 2479 |
+
"loss": 1.6875,
|
| 2480 |
+
"step": 1760
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 0.8816183816183816,
|
| 2484 |
+
"grad_norm": 0.1559111566296162,
|
| 2485 |
+
"learning_rate": 8.424511382285749e-07,
|
| 2486 |
+
"loss": 1.7169,
|
| 2487 |
+
"step": 1765
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 0.8841158841158842,
|
| 2491 |
+
"grad_norm": 0.15776692471495612,
|
| 2492 |
+
"learning_rate": 8.077611455065493e-07,
|
| 2493 |
+
"loss": 1.7276,
|
| 2494 |
+
"step": 1770
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 0.8866133866133866,
|
| 2498 |
+
"grad_norm": 0.15457624124158628,
|
| 2499 |
+
"learning_rate": 7.737704002100255e-07,
|
| 2500 |
+
"loss": 1.6813,
|
| 2501 |
+
"step": 1775
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.8891108891108891,
|
| 2505 |
+
"grad_norm": 0.1542971837033228,
|
| 2506 |
+
"learning_rate": 7.404814879921296e-07,
|
| 2507 |
+
"loss": 1.7033,
|
| 2508 |
+
"step": 1780
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.8916083916083916,
|
| 2512 |
+
"grad_norm": 0.15193831756852316,
|
| 2513 |
+
"learning_rate": 7.078969411180159e-07,
|
| 2514 |
+
"loss": 1.6647,
|
| 2515 |
+
"step": 1785
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.8941058941058941,
|
| 2519 |
+
"grad_norm": 0.15247095933972296,
|
| 2520 |
+
"learning_rate": 6.760192382722485e-07,
|
| 2521 |
+
"loss": 1.7022,
|
| 2522 |
+
"step": 1790
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 0.8966033966033966,
|
| 2526 |
+
"grad_norm": 0.15342741598547197,
|
| 2527 |
+
"learning_rate": 6.44850804370234e-07,
|
| 2528 |
+
"loss": 1.7011,
|
| 2529 |
+
"step": 1795
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.8991008991008991,
|
| 2533 |
+
"grad_norm": 0.15441584662597227,
|
| 2534 |
+
"learning_rate": 6.143940103737689e-07,
|
| 2535 |
+
"loss": 1.7223,
|
| 2536 |
+
"step": 1800
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 0.9015984015984015,
|
| 2540 |
+
"grad_norm": 0.15544328956589934,
|
| 2541 |
+
"learning_rate": 5.846511731106864e-07,
|
| 2542 |
+
"loss": 1.7176,
|
| 2543 |
+
"step": 1805
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.9040959040959041,
|
| 2547 |
+
"grad_norm": 0.15335771344533347,
|
| 2548 |
+
"learning_rate": 5.556245550986051e-07,
|
| 2549 |
+
"loss": 1.7273,
|
| 2550 |
+
"step": 1810
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.9065934065934066,
|
| 2554 |
+
"grad_norm": 0.15533326348626406,
|
| 2555 |
+
"learning_rate": 5.273163643728296e-07,
|
| 2556 |
+
"loss": 1.6976,
|
| 2557 |
+
"step": 1815
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.9090909090909091,
|
| 2561 |
+
"grad_norm": 0.1516521943992359,
|
| 2562 |
+
"learning_rate": 4.997287543183815e-07,
|
| 2563 |
+
"loss": 1.6956,
|
| 2564 |
+
"step": 1820
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.9115884115884116,
|
| 2568 |
+
"grad_norm": 0.15363897561993292,
|
| 2569 |
+
"learning_rate": 4.728638235062022e-07,
|
| 2570 |
+
"loss": 1.6807,
|
| 2571 |
+
"step": 1825
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 0.9140859140859141,
|
| 2575 |
+
"grad_norm": 0.15398450709114828,
|
| 2576 |
+
"learning_rate": 4.4672361553350307e-07,
|
| 2577 |
+
"loss": 1.7005,
|
| 2578 |
+
"step": 1830
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 0.9165834165834166,
|
| 2582 |
+
"grad_norm": 0.15239411554319313,
|
| 2583 |
+
"learning_rate": 4.213101188683155e-07,
|
| 2584 |
+
"loss": 1.6957,
|
| 2585 |
+
"step": 1835
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.919080919080919,
|
| 2589 |
+
"grad_norm": 0.15568622874699023,
|
| 2590 |
+
"learning_rate": 3.9662526669823954e-07,
|
| 2591 |
+
"loss": 1.7054,
|
| 2592 |
+
"step": 1840
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.9215784215784216,
|
| 2596 |
+
"grad_norm": 0.15485180274139432,
|
| 2597 |
+
"learning_rate": 3.7267093678336807e-07,
|
| 2598 |
+
"loss": 1.7272,
|
| 2599 |
+
"step": 1845
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.9240759240759241,
|
| 2603 |
+
"grad_norm": 0.15780700441318615,
|
| 2604 |
+
"learning_rate": 3.4944895131346355e-07,
|
| 2605 |
+
"loss": 1.6873,
|
| 2606 |
+
"step": 1850
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 0.9265734265734266,
|
| 2610 |
+
"grad_norm": 0.15379026482093777,
|
| 2611 |
+
"learning_rate": 3.2696107676933874e-07,
|
| 2612 |
+
"loss": 1.6611,
|
| 2613 |
+
"step": 1855
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 0.929070929070929,
|
| 2617 |
+
"grad_norm": 0.1548358187558438,
|
| 2618 |
+
"learning_rate": 3.052090237884808e-07,
|
| 2619 |
+
"loss": 1.7334,
|
| 2620 |
+
"step": 1860
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.9315684315684316,
|
| 2624 |
+
"grad_norm": 0.15404474230711981,
|
| 2625 |
+
"learning_rate": 2.841944470349256e-07,
|
| 2626 |
+
"loss": 1.7395,
|
| 2627 |
+
"step": 1865
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 0.9340659340659341,
|
| 2631 |
+
"grad_norm": 0.15245851102095165,
|
| 2632 |
+
"learning_rate": 2.6391894507339036e-07,
|
| 2633 |
+
"loss": 1.6845,
|
| 2634 |
+
"step": 1870
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 0.9365634365634365,
|
| 2638 |
+
"grad_norm": 0.15457664788214828,
|
| 2639 |
+
"learning_rate": 2.443840602476666e-07,
|
| 2640 |
+
"loss": 1.7197,
|
| 2641 |
+
"step": 1875
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.939060939060939,
|
| 2645 |
+
"grad_norm": 0.15619921990085028,
|
| 2646 |
+
"learning_rate": 2.2559127856330187e-07,
|
| 2647 |
+
"loss": 1.7138,
|
| 2648 |
+
"step": 1880
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 0.9415584415584416,
|
| 2652 |
+
"grad_norm": 0.15465495381686764,
|
| 2653 |
+
"learning_rate": 2.075420295745567e-07,
|
| 2654 |
+
"loss": 1.7034,
|
| 2655 |
+
"step": 1885
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.9440559440559441,
|
| 2659 |
+
"grad_norm": 0.15413725828913954,
|
| 2660 |
+
"learning_rate": 1.902376862756583e-07,
|
| 2661 |
+
"loss": 1.7132,
|
| 2662 |
+
"step": 1890
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.9465534465534465,
|
| 2666 |
+
"grad_norm": 0.1553239253397568,
|
| 2667 |
+
"learning_rate": 1.7367956499635963e-07,
|
| 2668 |
+
"loss": 1.7267,
|
| 2669 |
+
"step": 1895
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.949050949050949,
|
| 2673 |
+
"grad_norm": 0.15270107237353941,
|
| 2674 |
+
"learning_rate": 1.578689253018062e-07,
|
| 2675 |
+
"loss": 1.7124,
|
| 2676 |
+
"step": 1900
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.9515484515484516,
|
| 2680 |
+
"grad_norm": 0.15613663746353207,
|
| 2681 |
+
"learning_rate": 1.4280696989672383e-07,
|
| 2682 |
+
"loss": 1.6997,
|
| 2683 |
+
"step": 1905
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.954045954045954,
|
| 2687 |
+
"grad_norm": 0.15490386612035345,
|
| 2688 |
+
"learning_rate": 1.2849484453392624e-07,
|
| 2689 |
+
"loss": 1.7035,
|
| 2690 |
+
"step": 1910
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.9565434565434565,
|
| 2694 |
+
"grad_norm": 0.15328450998495585,
|
| 2695 |
+
"learning_rate": 1.1493363792716262e-07,
|
| 2696 |
+
"loss": 1.7039,
|
| 2697 |
+
"step": 1915
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.9590409590409591,
|
| 2701 |
+
"grad_norm": 0.15356789205414245,
|
| 2702 |
+
"learning_rate": 1.0212438166829375e-07,
|
| 2703 |
+
"loss": 1.7214,
|
| 2704 |
+
"step": 1920
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.9615384615384616,
|
| 2708 |
+
"grad_norm": 0.15318911423373824,
|
| 2709 |
+
"learning_rate": 9.006805014882825e-08,
|
| 2710 |
+
"loss": 1.6622,
|
| 2711 |
+
"step": 1925
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.964035964035964,
|
| 2715 |
+
"grad_norm": 0.15361081252860728,
|
| 2716 |
+
"learning_rate": 7.876556048579287e-08,
|
| 2717 |
+
"loss": 1.6955,
|
| 2718 |
+
"step": 1930
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.9665334665334665,
|
| 2722 |
+
"grad_norm": 0.15466222218440107,
|
| 2723 |
+
"learning_rate": 6.82177724519717e-08,
|
| 2724 |
+
"loss": 1.6808,
|
| 2725 |
+
"step": 1935
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.9690309690309691,
|
| 2729 |
+
"grad_norm": 0.1551486540401332,
|
| 2730 |
+
"learning_rate": 5.8425488410505107e-08,
|
| 2731 |
+
"loss": 1.7217,
|
| 2732 |
+
"step": 1940
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.9715284715284715,
|
| 2736 |
+
"grad_norm": 0.15270102301709607,
|
| 2737 |
+
"learning_rate": 4.9389453253850806e-08,
|
| 2738 |
+
"loss": 1.674,
|
| 2739 |
+
"step": 1945
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 0.974025974025974,
|
| 2743 |
+
"grad_norm": 0.15535594613189269,
|
| 2744 |
+
"learning_rate": 4.111035434712585e-08,
|
| 2745 |
+
"loss": 1.7045,
|
| 2746 |
+
"step": 1950
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 0.9765234765234765,
|
| 2750 |
+
"grad_norm": 0.15346069144811603,
|
| 2751 |
+
"learning_rate": 3.35888214758151e-08,
|
| 2752 |
+
"loss": 1.714,
|
| 2753 |
+
"step": 1955
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 0.9790209790209791,
|
| 2757 |
+
"grad_norm": 0.1535377723271678,
|
| 2758 |
+
"learning_rate": 2.682542679786071e-08,
|
| 2759 |
+
"loss": 1.7105,
|
| 2760 |
+
"step": 1960
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 0.9815184815184815,
|
| 2764 |
+
"grad_norm": 0.153091026718829,
|
| 2765 |
+
"learning_rate": 2.0820684800147983e-08,
|
| 2766 |
+
"loss": 1.6672,
|
| 2767 |
+
"step": 1965
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 0.984015984015984,
|
| 2771 |
+
"grad_norm": 0.15163180759905762,
|
| 2772 |
+
"learning_rate": 1.557505225936118e-08,
|
| 2773 |
+
"loss": 1.7388,
|
| 2774 |
+
"step": 1970
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 0.9865134865134865,
|
| 2778 |
+
"grad_norm": 0.15480094009362919,
|
| 2779 |
+
"learning_rate": 1.1088928207236837e-08,
|
| 2780 |
+
"loss": 1.6973,
|
| 2781 |
+
"step": 1975
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 0.989010989010989,
|
| 2785 |
+
"grad_norm": 0.15658829069949848,
|
| 2786 |
+
"learning_rate": 7.362653900215844e-09,
|
| 2787 |
+
"loss": 1.7533,
|
| 2788 |
+
"step": 1980
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 0.9915084915084915,
|
| 2792 |
+
"grad_norm": 0.15216671113258667,
|
| 2793 |
+
"learning_rate": 4.396512793475305e-09,
|
| 2794 |
+
"loss": 1.6998,
|
| 2795 |
+
"step": 1985
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 0.994005994005994,
|
| 2799 |
+
"grad_norm": 0.15475133859641596,
|
| 2800 |
+
"learning_rate": 2.1907305193757944e-09,
|
| 2801 |
+
"loss": 1.7031,
|
| 2802 |
+
"step": 1990
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 0.9965034965034965,
|
| 2806 |
+
"grad_norm": 0.15399579527373794,
|
| 2807 |
+
"learning_rate": 7.454748702895309e-10,
|
| 2808 |
+
"loss": 1.6964,
|
| 2809 |
+
"step": 1995
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 0.999000999000999,
|
| 2813 |
+
"grad_norm": 0.157143946266898,
|
| 2814 |
+
"learning_rate": 6.085578584280604e-11,
|
| 2815 |
+
"loss": 1.7337,
|
| 2816 |
+
"step": 2000
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 1.0,
|
| 2820 |
+
"eval_loss": 1.7156749963760376,
|
| 2821 |
+
"eval_runtime": 110.9741,
|
| 2822 |
+
"eval_samples_per_second": 127.76,
|
| 2823 |
+
"eval_steps_per_second": 2.0,
|
| 2824 |
+
"step": 2002
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 1.0,
|
| 2828 |
+
"step": 2002,
|
| 2829 |
+
"total_flos": 145100101386240.0,
|
| 2830 |
+
"train_loss": 1.7762367073948924,
|
| 2831 |
+
"train_runtime": 4136.0894,
|
| 2832 |
+
"train_samples_per_second": 30.978,
|
| 2833 |
+
"train_steps_per_second": 0.484
|
| 2834 |
+
}
|
| 2835 |
+
],
|
| 2836 |
+
"logging_steps": 5,
|
| 2837 |
+
"max_steps": 2002,
|
| 2838 |
+
"num_input_tokens_seen": 0,
|
| 2839 |
+
"num_train_epochs": 1,
|
| 2840 |
+
"save_steps": 100,
|
| 2841 |
+
"stateful_callbacks": {
|
| 2842 |
+
"TrainerControl": {
|
| 2843 |
+
"args": {
|
| 2844 |
+
"should_epoch_stop": false,
|
| 2845 |
+
"should_evaluate": false,
|
| 2846 |
+
"should_log": false,
|
| 2847 |
+
"should_save": true,
|
| 2848 |
+
"should_training_stop": true
|
| 2849 |
+
},
|
| 2850 |
+
"attributes": {}
|
| 2851 |
+
}
|
| 2852 |
+
},
|
| 2853 |
+
"total_flos": 145100101386240.0,
|
| 2854 |
+
"train_batch_size": 8,
|
| 2855 |
+
"trial_name": null,
|
| 2856 |
+
"trial_params": null
|
| 2857 |
+
}
|