Update README.md
Browse files
README.md
CHANGED
|
@@ -21,3 +21,14 @@ language:
|
|
| 21 |
This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
|
| 22 |
|
| 23 |
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
|
| 22 |
|
| 23 |
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
|
| 24 |
+
|
| 25 |
+
|
| 26 |
+
|
| 27 |
+
2024年9月、10月のニュース情報を継続事前学習したもの
|
| 28 |
+
epoch:3
|
| 29 |
+
r:128
|
| 30 |
+
lora_alpha:512
|
| 31 |
+
lr:3e-4
|
| 32 |
+
embedding_lr: 3e-5
|
| 33 |
+
|
| 34 |
+
狙い: r、lora_alphaを大きくして、SFTによる記憶忘却に耐えられるようにCPTによるベースモデルへの知識埋め込みの影響を大きくする
|