Update README.md
Browse files
README.md
CHANGED
|
@@ -20,9 +20,8 @@ tags:
|
|
| 20 |
C3TR-AdapterのVersion3を公開しました。
|
| 21 |
Version 3 of C3TR-Adapter has been released.
|
| 22 |
|
| 23 |
-
|
| 24 |
-
|
| 25 |
-
|
| 26 |
|
| 27 |
## 2024.05.17
|
| 28 |
[C3TR-AdapterのVersion2](https://huggingface.co/webbigdata/C3TR-Adapter/tree/version2)を公開しました。
|
|
@@ -57,8 +56,8 @@ For instructions on how to run the translation benchmark and other benchmark res
|
|
| 57 |
GoogleのウェブサービスColabを使うと無料でC3TR-Adapterを試す事が出来ます。リンク先でOpen In Colabボタンを押して起動してください。
|
| 58 |
You can try C3TR-Adapter for free using Google's web service Colab. Please press the Open In Colab button on the link to activate it.
|
| 59 |
|
| 60 |
-
- [動作確認用の簡単なサンプル(A simple sample to check the operation)](https://github.com/webbigdata-jp/python_sample/blob/main/
|
| 61 |
-
- [テキストファイルを一括で日英・英日翻訳するサンプル(Sample of batch translation of text files)](https://github.com/webbigdata-jp/python_sample/blob/main/
|
| 62 |
- [GPUがない環境でも動かす事ができるgguf版(A gguf version that can be run in environments without a GPU)](https://huggingface.co/webbigdata/C3TR-Adapter_gguf)
|
| 63 |
|
| 64 |
### モデルの動かし方(How to use Model)
|
|
@@ -78,8 +77,8 @@ If you want to run it on your own local computer, you will need at least approxi
|
|
| 78 |
# example for windows user with CUDA 12.1.
|
| 79 |
# pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121
|
| 80 |
|
| 81 |
-
# Gemmaは最新のライブラリでなくては動かないので、以下のVersionに更新してください
|
| 82 |
-
# Gemma will not work without the latest library, so please update to the following version
|
| 83 |
pip install transformers==4.42.3
|
| 84 |
pip install peft==0.11.1
|
| 85 |
pip install bitsandbytes==0.43.1
|
|
@@ -96,7 +95,6 @@ from peft import PeftModel
|
|
| 96 |
model_id = "unsloth/gemma-2-9b-it-bnb-4bit"
|
| 97 |
peft_model_id = "webbigdata/C3TR-Adapter"
|
| 98 |
|
| 99 |
-
# デバイスがbfloat16をサポートしているかどうかを確認
|
| 100 |
if torch.cuda.is_available() and torch.cuda.get_device_capability(0)[0] >= 8:
|
| 101 |
dtype = torch.bfloat16
|
| 102 |
else:
|
|
@@ -475,7 +473,7 @@ Since today's meeting is very important, please arrive on time.
|
|
| 475 |
You can specify the first person and ending for your character.
|
| 476 |
This feature is still highly experimental and incomplete at this time.
|
| 477 |
|
| 478 |
-
映画「her」より一場面(A scene from the movie "
|
| 479 |
```
|
| 480 |
<start_of_turn>### Instruction:
|
| 481 |
Translate English to Japanese.
|
|
|
|
| 20 |
C3TR-AdapterのVersion3を公開しました。
|
| 21 |
Version 3 of C3TR-Adapter has been released.
|
| 22 |
|
| 23 |
+
version3では4つのベンチマークのうち、1つでgpt4 turboを上回るという大幅な性能底上げが達成されています。
|
| 24 |
+
Version 3 achieved a significant performance boost, beating GPT4 Turbo in one of the four benchmarks.
|
|
|
|
| 25 |
|
| 26 |
## 2024.05.17
|
| 27 |
[C3TR-AdapterのVersion2](https://huggingface.co/webbigdata/C3TR-Adapter/tree/version2)を公開しました。
|
|
|
|
| 56 |
GoogleのウェブサービスColabを使うと無料でC3TR-Adapterを試す事が出来ます。リンク先でOpen In Colabボタンを押して起動してください。
|
| 57 |
You can try C3TR-Adapter for free using Google's web service Colab. Please press the Open In Colab button on the link to activate it.
|
| 58 |
|
| 59 |
+
- [動作確認用の簡単なサンプル(A simple sample to check the operation)](https://github.com/webbigdata-jp/python_sample/blob/main/C3TR_Adapter_v3_Japanese_English_Translation_sample_code.ipynb)
|
| 60 |
+
- [テキストファイルを一括で日英・英日翻訳するサンプル(Sample of batch translation of text files)](https://github.com/webbigdata-jp/python_sample/blob/main/C3TR_Adapter_v3_batch_translation_sample.ipynb)
|
| 61 |
- [GPUがない環境でも動かす事ができるgguf版(A gguf version that can be run in environments without a GPU)](https://huggingface.co/webbigdata/C3TR-Adapter_gguf)
|
| 62 |
|
| 63 |
### モデルの動かし方(How to use Model)
|
|
|
|
| 77 |
# example for windows user with CUDA 12.1.
|
| 78 |
# pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121
|
| 79 |
|
| 80 |
+
# Gemma 2は最新のライブラリでなくては動かないので、以下のVersionに更新してください
|
| 81 |
+
# Gemma 2 will not work without the latest library, so please update to the following version
|
| 82 |
pip install transformers==4.42.3
|
| 83 |
pip install peft==0.11.1
|
| 84 |
pip install bitsandbytes==0.43.1
|
|
|
|
| 95 |
model_id = "unsloth/gemma-2-9b-it-bnb-4bit"
|
| 96 |
peft_model_id = "webbigdata/C3TR-Adapter"
|
| 97 |
|
|
|
|
| 98 |
if torch.cuda.is_available() and torch.cuda.get_device_capability(0)[0] >= 8:
|
| 99 |
dtype = torch.bfloat16
|
| 100 |
else:
|
|
|
|
| 473 |
You can specify the first person and ending for your character.
|
| 474 |
This feature is still highly experimental and incomplete at this time.
|
| 475 |
|
| 476 |
+
映画「her」より一場面(A scene from the movie "her")
|
| 477 |
```
|
| 478 |
<start_of_turn>### Instruction:
|
| 479 |
Translate English to Japanese.
|