Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
datasets:
|
| 3 |
+
- IlyaGusev/ru_turbo_alpaca
|
| 4 |
+
- IlyaGusev/ru_turbo_saiga
|
| 5 |
+
- IlyaGusev/ru_sharegpt_cleaned
|
| 6 |
+
- IlyaGusev/oasst1_ru_main_branch
|
| 7 |
+
- IlyaGusev/ru_turbo_alpaca_evol_instruct
|
| 8 |
+
- lksy/ru_instruct_gpt4
|
| 9 |
+
language:
|
| 10 |
+
- ru
|
| 11 |
+
inference: false
|
| 12 |
+
pipeline_tag: conversational
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
Llama.cpp compatible versions of an original [13B model](https://huggingface.co/IlyaGusev/saiga2_13b_lora).
|
| 16 |
+
|
| 17 |
+
* Download one of the versions, for example `ggml-model-q4_1.bin`.
|
| 18 |
+
* Download [interact_llamacpp.py](https://raw.githubusercontent.com/IlyaGusev/rulm/master/self_instruct/src/interact_llamacpp.py)
|
| 19 |
+
|
| 20 |
+
How to run:
|
| 21 |
+
```
|
| 22 |
+
sudo apt-get install git-lfs
|
| 23 |
+
pip install llama-cpp-python fire
|
| 24 |
+
|
| 25 |
+
python3 interact_llamacpp.py ggml-model-q4_1.bin
|
| 26 |
+
```
|
| 27 |
+
|
| 28 |
+
System requirements:
|
| 29 |
+
* 18GB RAM for q8_0
|
| 30 |
+
* 13GB RAM for q4_1
|