38 lines
767 B
Markdown
38 lines
767 B
Markdown
|
---
|
|||
|
|
library_name: transformers
|
|||
|
|
language:
|
|||
|
|
- ru
|
|||
|
|
- en
|
|||
|
|
pipeline_tag: text-generation
|
|||
|
|
license: other
|
|||
|
|
license_name: apache-2.0
|
|||
|
|
license_link: https://huggingface.co/MTSAIR/Cotype-Nano-GGUF/blob/main/Apache%20License%20MTS%20AI.docx
|
|||
|
|
---
|
|||
|
|
|
|||
|
|
# Cotype Nano GGUF🤖
|
|||
|
|
|
|||
|
|
Cotype-Nano-GGUF– это легковесная LLM, которая может запускаться даже на мобильных девайсах
|
|||
|
|
|
|||
|
|
|
|||
|
|
|
|||
|
|
Cotype-Nano-GGUF is a lightweight LLM that can run even on mobile devices
|
|||
|
|
|
|||
|
|
### Inference
|
|||
|
|
|
|||
|
|
```python
|
|||
|
|
from llama_cpp import Llama
|
|||
|
|
|
|||
|
|
llm = Llama.from_pretrained(
|
|||
|
|
repo_id="MTSAIR/Cotype-Nano-GGUF",
|
|||
|
|
filename="cotype_nano_8bit.gguf",
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
llm.create_chat_completion(
|
|||
|
|
messages = [
|
|||
|
|
{
|
|||
|
|
"role": "user",
|
|||
|
|
"content": "What is the capital of France?"
|
|||
|
|
}
|
|||
|
|
]
|
|||
|
|
)
|
|||
|
|
```
|