File size: 956 Bytes
b674524 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 |
---
datasets:
- IlyaGusev/ru_turbo_alpaca
- IlyaGusev/ru_turbo_saiga
- IlyaGusev/ru_sharegpt_cleaned
- IlyaGusev/oasst1_ru_main_branch
- IlyaGusev/ru_turbo_alpaca_evol_instruct
- lksy/ru_instruct_gpt4
language:
- ru
inference: false
pipeline_tag: conversational
license: llama2
---
Llama.cpp compatible versions of an original [7B model](https://huggingface.co/IlyaGusev/saiga2_7b_lora).
Download one of the versions, for example `model-q4_K.gguf`.
```
wget https://huggingface.co/IlyaGusev/saiga2_7b_gguf/resolve/main/model-q4_K.gguf
```
Download [interact_llamacpp.py](https://raw.githubusercontent.com/IlyaGusev/rulm/master/self_instruct/src/interact_llamacpp.py)
```
wget https://raw.githubusercontent.com/IlyaGusev/rulm/master/self_instruct/src/interact_llamacpp.py
```
How to run:
```
pip install llama-cpp-python fire
python3 interact_llamacpp.py model-q4_K.gguf
```
System requirements:
* 10GB RAM for q8_0 and less for smaller quantizations
|