mradermacher commited on
Commit
2b7c445
1 Parent(s): 91ba49a

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +4 -0
README.md CHANGED
@@ -33,7 +33,11 @@ more details, including on how to concatenate multi-part files.
33
  |:-----|:-----|--------:|:------|
34
  | [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-hf-GGUF/resolve/main/CodeLlama-13b-hf.Q2_K.gguf) | Q2_K | 5.0 | |
35
  | [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-hf-GGUF/resolve/main/CodeLlama-13b-hf.IQ3_S.gguf) | IQ3_S | 5.8 | beats Q3_K* |
 
 
 
36
  | [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-hf-GGUF/resolve/main/CodeLlama-13b-hf.Q4_K_S.gguf) | Q4_K_S | 7.5 | fast, recommended |
 
37
  | [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-hf-GGUF/resolve/main/CodeLlama-13b-hf.Q8_0.gguf) | Q8_0 | 13.9 | fast, best quality |
38
 
39
  Here is a handy graph by ikawrakow comparing some lower-quality quant
 
33
  |:-----|:-----|--------:|:------|
34
  | [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-hf-GGUF/resolve/main/CodeLlama-13b-hf.Q2_K.gguf) | Q2_K | 5.0 | |
35
  | [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-hf-GGUF/resolve/main/CodeLlama-13b-hf.IQ3_S.gguf) | IQ3_S | 5.8 | beats Q3_K* |
36
+ | [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-hf-GGUF/resolve/main/CodeLlama-13b-hf.Q3_K_S.gguf) | Q3_K_S | 5.8 | |
37
+ | [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-hf-GGUF/resolve/main/CodeLlama-13b-hf.IQ3_M.gguf) | IQ3_M | 6.1 | |
38
+ | [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-hf-GGUF/resolve/main/CodeLlama-13b-hf.Q3_K_M.gguf) | Q3_K_M | 6.4 | lower quality |
39
  | [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-hf-GGUF/resolve/main/CodeLlama-13b-hf.Q4_K_S.gguf) | Q4_K_S | 7.5 | fast, recommended |
40
+ | [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-hf-GGUF/resolve/main/CodeLlama-13b-hf.Q6_K.gguf) | Q6_K | 10.8 | very good quality |
41
  | [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-hf-GGUF/resolve/main/CodeLlama-13b-hf.Q8_0.gguf) | Q8_0 | 13.9 | fast, best quality |
42
 
43
  Here is a handy graph by ikawrakow comparing some lower-quality quant