Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,32 @@
|
|
| 1 |
---
|
| 2 |
license: bsd
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: bsd
|
| 3 |
---
|
| 4 |
+
|
| 5 |
+
Converted INT8/INT4 files for [fastllm](https://github.com/ztxz16/fastllm) with [baichuan2-13b-chat](https://huggingface.co/baichuan-inc/Baichuan2-13B-Chat)
|
| 6 |
+
|
| 7 |
+
Directly download from Baidu Netdisk:
|
| 8 |
+
|
| 9 |
+
Link：https://pan.baidu.com/s/1Xsiif_1VzDyWFei1u5oJcA
|
| 10 |
+
Code：wxbo
|
| 11 |
+
|
| 12 |
+
Updated time: 2023/09/11
|
| 13 |
+
|
| 14 |
+
|
| 15 |
+
```
|
| 16 |
+
|
| 17 |
+
baichuan2-13b-chat-int8.flm:
|
| 18 |
+
|
| 19 |
+
|===============================+======================+======================|
|
| 20 |
+
| 0 NVIDIA GeForce ... Off | 00000000:05:00.0 Off | N/A |
|
| 21 |
+
| 31% 36C P8 28W / 250W | 15420MiB / 22528MiB | 0% Default |
|
| 22 |
+
+-------------------------------+----------------------+----------------------+
|
| 23 |
+
|
| 24 |
+
```
|
| 25 |
+
|
| 26 |
+
```python
|
| 27 |
+
from fastllm_pytools import llm
|
| 28 |
+
model = llm.model("baichuan2-13b-chat-int8.flm")
|
| 29 |
+
for response in model.stream_response("介绍一下南京"):
|
| 30 |
+
print(response, flush = True, end = "")
|
| 31 |
+
```
|
| 32 |
+
> Note: please use the lastest version of FastLLM (no later than 2023/09/11 main branch)
|