Hugging Face's logo

mradermacher
/
RoGemma-7b-Instruct-2024-06-28-GGUF

Transformers
GGUF
Romanian
conversational
Model card Files Community
mradermacher commited on
Commit
51cca7a
·
verified ·
1 Parent(s): e8b3cad

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -0
README.md CHANGED
@@ -41,7 +41,14 @@ more details, including on how to concatenate multi-part files.
41
  | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.Q3_K_S.gguf) | Q3_K_S | 4.1 | |
42
  | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.Q3_K_M.gguf) | Q3_K_M | 4.5 | lower quality |
43
  | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.Q3_K_L.gguf) | Q3_K_L | 4.8 | |
 
44
  | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.Q4_K_S.gguf) | Q4_K_S | 5.1 | fast, recommended |
 
 
 
 
 
 
45
 
46
  Here is a handy graph by ikawrakow comparing some lower-quality quant
47
  types (lower is better):
 
41
  | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.Q3_K_S.gguf) | Q3_K_S | 4.1 | |
42
  | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.Q3_K_M.gguf) | Q3_K_M | 4.5 | lower quality |
43
  | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.Q3_K_L.gguf) | Q3_K_L | 4.8 | |
44
+ | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.IQ4_XS.gguf) | IQ4_XS | 4.9 | |
45
  | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.Q4_K_S.gguf) | Q4_K_S | 5.1 | fast, recommended |
46
+ | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.Q4_K_M.gguf) | Q4_K_M | 5.4 | fast, recommended |
47
+ | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.Q5_K_S.gguf) | Q5_K_S | 6.1 | |
48
+ | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.Q5_K_M.gguf) | Q5_K_M | 6.2 | |
49
+ | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.Q6_K.gguf) | Q6_K | 7.1 | very good quality |
50
+ | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.Q8_0.gguf) | Q8_0 | 9.2 | fast, best quality |
51
+ | [GGUF](https://huggingface.co/mradermacher/RoGemma-7b-Instruct-2024-06-28-GGUF/resolve/main/RoGemma-7b-Instruct-2024-06-28.f16.gguf) | f16 | 17.2 | 16 bpw, overkill |
52
 
53
  Here is a handy graph by ikawrakow comparing some lower-quality quant
54
  types (lower is better):