Commit
•
3256fec
1
Parent(s):
f0ab64e
Update README.md
Browse files
README.md
CHANGED
|
@@ -15,6 +15,9 @@ tags:
|
|
| 15 |
- orpo
|
| 16 |
---
|
| 17 |
|
|
|
|
|
|
|
|
|
|
| 18 |
# 1. Introduction
|
| 19 |
|
| 20 |
This is the first Chinese chat model specifically fine-tuned for Chinese through ORPO [1] based on the [Meta-Llama-3-8B-Instruct model](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct).
|
|
@@ -97,7 +100,7 @@ model = AutoModelForCausalLM.from_pretrained(
|
|
| 97 |
)
|
| 98 |
|
| 99 |
messages = [
|
| 100 |
-
{"role": "system", "content": "You are a helpful assistant."},
|
| 101 |
{"role": "user", "content": "介绍一下你自己"},
|
| 102 |
]
|
| 103 |
|
|
|
|
| 15 |
- orpo
|
| 16 |
---
|
| 17 |
|
| 18 |
+
# Updates:
|
| 19 |
+
- 🔥 We provide official 8bit-quantized GGUF versions of Llama3-8B-Chinese-Chat in [shenzhi-wang/Llama3-8B-Chinese-Chat-GGUF-8bit](https://huggingface.co/shenzhi-wang/Llama3-8B-Chinese-Chat-GGUF-8bit)!
|
| 20 |
+
|
| 21 |
# 1. Introduction
|
| 22 |
|
| 23 |
This is the first Chinese chat model specifically fine-tuned for Chinese through ORPO [1] based on the [Meta-Llama-3-8B-Instruct model](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct).
|
|
|
|
| 100 |
)
|
| 101 |
|
| 102 |
messages = [
|
| 103 |
+
{"role": "system", "content": "You are Llama3-8B-Chinese-Chat, which is finetuned on Llama3-8B-Instruct with Chinese-English mixed data by the ORPO alignment algorithm, developed by Shenzhi Wang (王慎执 in Chinese). You are a helpful assistant."},
|
| 104 |
{"role": "user", "content": "介绍一下你自己"},
|
| 105 |
]
|
| 106 |
|