mradermacher
commited on
Commit
•
f8d7e15
1
Parent(s):
c5def3c
auto-patch README.md
Browse files
README.md
CHANGED
@@ -175,7 +175,7 @@ tags:
|
|
175 |
static quants of https://huggingface.co/huihui-ai/CodeLlama-13b-Instruct-hf-abliterated
|
176 |
|
177 |
<!-- provided-files -->
|
178 |
-
weighted/imatrix quants
|
179 |
## Usage
|
180 |
|
181 |
If you are unsure how to use GGUF files, refer to one of [TheBloke's
|
@@ -191,7 +191,14 @@ more details, including on how to concatenate multi-part files.
|
|
191 |
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q2_K.gguf) | Q2_K | 5.0 | |
|
192 |
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q3_K_S.gguf) | Q3_K_S | 5.8 | |
|
193 |
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q3_K_M.gguf) | Q3_K_M | 6.4 | lower quality |
|
|
|
|
|
194 |
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q4_K_S.gguf) | Q4_K_S | 7.5 | fast, recommended |
|
|
|
|
|
|
|
|
|
|
|
195 |
|
196 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
197 |
types (lower is better):
|
|
|
175 |
static quants of https://huggingface.co/huihui-ai/CodeLlama-13b-Instruct-hf-abliterated
|
176 |
|
177 |
<!-- provided-files -->
|
178 |
+
weighted/imatrix quants are available at https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-i1-GGUF
|
179 |
## Usage
|
180 |
|
181 |
If you are unsure how to use GGUF files, refer to one of [TheBloke's
|
|
|
191 |
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q2_K.gguf) | Q2_K | 5.0 | |
|
192 |
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q3_K_S.gguf) | Q3_K_S | 5.8 | |
|
193 |
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q3_K_M.gguf) | Q3_K_M | 6.4 | lower quality |
|
194 |
+
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q3_K_L.gguf) | Q3_K_L | 7.0 | |
|
195 |
+
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.IQ4_XS.gguf) | IQ4_XS | 7.1 | |
|
196 |
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q4_K_S.gguf) | Q4_K_S | 7.5 | fast, recommended |
|
197 |
+
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q4_K_M.gguf) | Q4_K_M | 8.0 | fast, recommended |
|
198 |
+
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q5_K_S.gguf) | Q5_K_S | 9.1 | |
|
199 |
+
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q5_K_M.gguf) | Q5_K_M | 9.3 | |
|
200 |
+
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q6_K.gguf) | Q6_K | 10.8 | very good quality |
|
201 |
+
| [GGUF](https://huggingface.co/mradermacher/CodeLlama-13b-Instruct-hf-abliterated-GGUF/resolve/main/CodeLlama-13b-Instruct-hf-abliterated.Q8_0.gguf) | Q8_0 | 13.9 | fast, best quality |
|
202 |
|
203 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
204 |
types (lower is better):
|