auto-patch README.md
Browse files
README.md
CHANGED
|
@@ -35,6 +35,18 @@ more details, including on how to concatenate multi-part files.
|
|
| 35 |
|:-----|:-----|--------:|:------|
|
| 36 |
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.mmproj-Q8_0.gguf) | mmproj-Q8_0 | 0.7 | multi-modal supplement |
|
| 37 |
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.mmproj-f16.gguf) | mmproj-f16 | 1.0 | multi-modal supplement |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
|
| 39 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
| 40 |
types (lower is better):
|
|
|
|
| 35 |
|:-----|:-----|--------:|:------|
|
| 36 |
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.mmproj-Q8_0.gguf) | mmproj-Q8_0 | 0.7 | multi-modal supplement |
|
| 37 |
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.mmproj-f16.gguf) | mmproj-f16 | 1.0 | multi-modal supplement |
|
| 38 |
+
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.Q2_K.gguf) | Q2_K | 2.0 | |
|
| 39 |
+
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.Q3_K_S.gguf) | Q3_K_S | 2.3 | |
|
| 40 |
+
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.Q3_K_M.gguf) | Q3_K_M | 2.5 | lower quality |
|
| 41 |
+
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.Q3_K_L.gguf) | Q3_K_L | 2.6 | |
|
| 42 |
+
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.IQ4_XS.gguf) | IQ4_XS | 2.7 | |
|
| 43 |
+
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.Q4_K_S.gguf) | Q4_K_S | 2.9 | fast, recommended |
|
| 44 |
+
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.Q4_K_M.gguf) | Q4_K_M | 3.0 | fast, recommended |
|
| 45 |
+
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.Q5_K_S.gguf) | Q5_K_S | 3.3 | |
|
| 46 |
+
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.Q5_K_M.gguf) | Q5_K_M | 3.4 | |
|
| 47 |
+
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.Q6_K.gguf) | Q6_K | 3.8 | very good quality |
|
| 48 |
+
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.Q8_0.gguf) | Q8_0 | 4.9 | fast, best quality |
|
| 49 |
+
| [GGUF](https://huggingface.co/mradermacher/Medra4b-abliterated-GGUF/resolve/main/Medra4b-abliterated.f16.gguf) | f16 | 9.2 | 16 bpw, overkill |
|
| 50 |
|
| 51 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
| 52 |
types (lower is better):
|