Downloads · 30 days
8
2% of all-time downloads
rugpt3large_for_qna_250k1
ERmak1581/rugpt3large_for_qna_250k1
rugpt3large_for_qna_250k1 is a text generation model from ERmak1581. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as mit.
The ai-forever/rugpt3largebasedongpt2 based model was fine tuned for Question-Answer tasks in Russian.
All-time downloads
489
Public
Parameters
760M
3 GB on disk
Likes
1
Public
Repo makeup
Click a slice to open those files.
.safetensors3 GB · 100%
At a glance
- Task
- Text Generation
- Library
- transformers
- License
- mit
- Model type
- gpt2
- Access
- Public
- Created
- May 2, 2024
- Updated
- May 2, 2024
- SHA
- 45c3ef9e
Model card
From the Hugging Face model README
The ai-forever/rugpt3large_based_on_gpt2 based model was fine tuned for Question-Answer tasks in Russian.
Версия: датасет 250тыс. строк, 1-ая эпоха. В дальнейшем будут появлятся другие модели.
Качество ответа: среднее
Формат запроса: <s> [user] Запрос [assistant] ... </s>
Пример использования:
from transformers import GPT2Tokenizer, GPT2LMHeadModel
model = GPT2LMHeadModel.from_pretrained("ERmak1581/rugpt3large_for_qna_250k1")
tokenizer = GPT2Tokenizer.from_pretrained("ERmak1581/rugpt3large_for_qna_250k1")
print(tokenizer.decode(model.generate(
tokenizer.encode('<s> [user] Почему небо синее? [assistant]',
return_tensors="pt"),
max_new_tokens=100, no_repeat_ngram_size=2, temperature=0.7, do_sample=True)[0]))
- Task
- Text Generation
- Library
- transformers
- Type
- gpt2
- License
- mit
- Languages
- ru
- Created
- May 2, 2024
- Updated
- May 2, 2024