JosephusCheung
commited on
Commit
·
2a97280
1
Parent(s):
eb6f7f8
Update README.md
Browse files
README.md
CHANGED
@@ -50,6 +50,9 @@ GPT2Tokenizer fixed by [Kerfuffle](https://github.com/KerfuffleV2) on [https://g
|
|
50 |
|
51 |
Thanks TheBloke for GGUF quants: [https://huggingface.co/TheBloke/CausalLM-14B-GGUF](https://huggingface.co/TheBloke/CausalLM-14B-GGUF)
|
52 |
|
|
|
|
|
|
|
53 |
|
54 |
# Read Me:
|
55 |
|
@@ -147,6 +150,10 @@ GPT2Tokenizer 支持由 [Kerfuffle](https://github.com/KerfuffleV2) 修复于 [h
|
|
147 |
|
148 |
感谢 TheBloke 制作 GGUF 版本量化模型: [https://huggingface.co/TheBloke/CausalLM-14B-GGUF](https://huggingface.co/TheBloke/CausalLM-14B-GGUF)
|
149 |
|
|
|
|
|
|
|
|
|
150 |
## 请读我:
|
151 |
|
152 |
另请参阅[7B版本](https://huggingface.co/CausalLM/7B)
|
|
|
50 |
|
51 |
Thanks TheBloke for GGUF quants: [https://huggingface.co/TheBloke/CausalLM-14B-GGUF](https://huggingface.co/TheBloke/CausalLM-14B-GGUF)
|
52 |
|
53 |
+
**Caution:** Unofficial GPTQ and AWQ models may have issues as they use Wikitext for calibration, while this model has undergone considerable training on a synthesized Wikipedia conversation dataset.
|
54 |
+
|
55 |
+
It is not recommended to use any form of quantization, but rather to use smaller-sized models, as the 7B and 14B versions have high consistency. However, if you do use model quantization, please use GGUF.
|
56 |
|
57 |
# Read Me:
|
58 |
|
|
|
150 |
|
151 |
感谢 TheBloke 制作 GGUF 版本量化模型: [https://huggingface.co/TheBloke/CausalLM-14B-GGUF](https://huggingface.co/TheBloke/CausalLM-14B-GGUF)
|
152 |
|
153 |
+
**注意:** 非官方 GPTQ 和 AWQ 模型可能存在问题,因为它们使用 Wikitext 进行校准,而该模型已经在合成的 Wikipedia 对话数据集上经过了大量的训练。
|
154 |
+
|
155 |
+
不建议使用任何形式的量化,而是使用较小尺寸的模型,因为7B和14B版本具有较高的一致性。 但是,如果您确实使用模型量化,请使用 GGUF。
|
156 |
+
|
157 |
## 请读我:
|
158 |
|
159 |
另请参阅[7B版本](https://huggingface.co/CausalLM/7B)
|