JosephusCheung commited on
Commit
f1f8b2c
·
1 Parent(s): 2a4a0cb

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +8 -0
README.md CHANGED
@@ -45,6 +45,10 @@ GPT2Tokenizer fixed by [Kerfuffle](https://github.com/KerfuffleV2) on [https://g
45
 
46
  Thanks TheBloke for GGUF quants: [https://huggingface.co/TheBloke/CausalLM-7B-GGUF](https://huggingface.co/TheBloke/CausalLM-7B-GGUF)
47
 
 
 
 
 
48
  ## Read Me:
49
 
50
  Also see [14B Version](https://huggingface.co/CausalLM/14B)
@@ -115,6 +119,10 @@ GPT2Tokenizer 支持由 [Kerfuffle](https://github.com/KerfuffleV2) 修复于 [h
115
 
116
  感谢 TheBloke 制作 GGUF 版本量化模型: [https://huggingface.co/TheBloke/CausalLM-7B-GGUF](https://huggingface.co/TheBloke/CausalLM-7B-GGUF)
117
 
 
 
 
 
118
  ## 请读我:
119
 
120
  另请参阅[14B版本](https://huggingface.co/CausalLM/14B)
 
45
 
46
  Thanks TheBloke for GGUF quants: [https://huggingface.co/TheBloke/CausalLM-7B-GGUF](https://huggingface.co/TheBloke/CausalLM-7B-GGUF)
47
 
48
+ **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.
49
+
50
+ 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.
51
+
52
  ## Read Me:
53
 
54
  Also see [14B Version](https://huggingface.co/CausalLM/14B)
 
119
 
120
  感谢 TheBloke 制作 GGUF 版本量化模型: [https://huggingface.co/TheBloke/CausalLM-7B-GGUF](https://huggingface.co/TheBloke/CausalLM-7B-GGUF)
121
 
122
+ **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。
123
+
124
+ 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。
125
+
126
  ## 请读我:
127
 
128
  另请参阅[14B版本](https://huggingface.co/CausalLM/14B)