Update README.md
Browse files
README.md
CHANGED
|
@@ -50,13 +50,14 @@ https://weblab.t.u-tokyo.ac.jp/lecture/course-list/large-language-model/
|
|
| 50 |
|
| 51 |
# -*- coding: utf-8 -*-
|
| 52 |
```
|
|
|
|
| 53 |
llm-jp-3-13b-finetune2.ipynb
|
| 54 |
|
| 55 |
Automatically generated by Colab.
|
| 56 |
|
| 57 |
Original file is located at
|
| 58 |
https://colab.research.google.com/drive/1TLF_HtVz6ng9ZAWe7lHy59kiHBg3_3y0
|
| 59 |
-
|
| 60 |
|
| 61 |
!pip uninstall unsloth -y
|
| 62 |
!pip install --upgrade --no-cache-dir "unsloth[colab-new] @ git+https://github.com/unslothai/unsloth.git"
|
|
@@ -127,9 +128,9 @@ prompt = """### 指示
|
|
| 127 |
|
| 128 |
|
| 129 |
|
| 130 |
-
|
| 131 |
formatting_prompts_func: 各データをプロンプトに合わせた形式に合わせる
|
| 132 |
-
|
| 133 |
EOS_TOKEN = tokenizer.eos_token # トークナイザーのEOSトークン（文末トークン）
|
| 134 |
def formatting_prompts_func(examples):
|
| 135 |
input = examples["text"] # 入力データ
|
|
@@ -156,7 +157,7 @@ output.disable_custom_widget_manager()
|
|
| 156 |
|
| 157 |
print(dataset["test"]["formatted_text"][3])
|
| 158 |
|
| 159 |
-
|
| 160 |
training_arguments: 学習の設定
|
| 161 |
|
| 162 |
- output_dir:
|
|
@@ -215,7 +216,7 @@ training_arguments: 学習の設定
|
|
| 215 |
|
| 216 |
- report_to:
|
| 217 |
- ログの送信先 ("wandb"/"tensorboard"など)
|
| 218 |
-
|
| 219 |
from trl import SFTTrainer
|
| 220 |
from transformers import TrainingArguments
|
| 221 |
from unsloth import is_bfloat16_supported
|
|
|
|
| 50 |
|
| 51 |
# -*- coding: utf-8 -*-
|
| 52 |
```
|
| 53 |
+
"""
|
| 54 |
llm-jp-3-13b-finetune2.ipynb
|
| 55 |
|
| 56 |
Automatically generated by Colab.
|
| 57 |
|
| 58 |
Original file is located at
|
| 59 |
https://colab.research.google.com/drive/1TLF_HtVz6ng9ZAWe7lHy59kiHBg3_3y0
|
| 60 |
+
"""
|
| 61 |
|
| 62 |
!pip uninstall unsloth -y
|
| 63 |
!pip install --upgrade --no-cache-dir "unsloth[colab-new] @ git+https://github.com/unslothai/unsloth.git"
|
|
|
|
| 128 |
|
| 129 |
|
| 130 |
|
| 131 |
+
"""
|
| 132 |
formatting_prompts_func: 各データをプロンプトに合わせた形式に合わせる
|
| 133 |
+
"""
|
| 134 |
EOS_TOKEN = tokenizer.eos_token # トークナイザーのEOSトークン（文末トークン）
|
| 135 |
def formatting_prompts_func(examples):
|
| 136 |
input = examples["text"] # 入力データ
|
|
|
|
| 157 |
|
| 158 |
print(dataset["test"]["formatted_text"][3])
|
| 159 |
|
| 160 |
+
"""
|
| 161 |
training_arguments: 学習の設定
|
| 162 |
|
| 163 |
- output_dir:
|
|
|
|
| 216 |
|
| 217 |
- report_to:
|
| 218 |
- ログの送信先 ("wandb"/"tensorboard"など)
|
| 219 |
+
"""
|
| 220 |
from trl import SFTTrainer
|
| 221 |
from transformers import TrainingArguments
|
| 222 |
from unsloth import is_bfloat16_supported
|