auto-patch README.md
Browse files
README.md
CHANGED
@@ -34,7 +34,13 @@ more details, including on how to concatenate multi-part files.
|
|
34 |
| Link | Type | Size/GB | Notes |
|
35 |
|:-----|:-----|--------:|:------|
|
36 |
| [GGUF](https://huggingface.co/mradermacher/Swallow-MoE-2x13B-v0.1-GGUF/resolve/main/Swallow-MoE-2x13B-v0.1.Q2_K.gguf) | Q2_K | 8.1 | |
|
|
|
|
|
|
|
37 |
| [GGUF](https://huggingface.co/mradermacher/Swallow-MoE-2x13B-v0.1-GGUF/resolve/main/Swallow-MoE-2x13B-v0.1.Q4_K_S.gguf) | Q4_K_S | 12.4 | fast, recommended |
|
|
|
|
|
|
|
38 |
|
39 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
40 |
types (lower is better):
|
|
|
34 |
| Link | Type | Size/GB | Notes |
|
35 |
|:-----|:-----|--------:|:------|
|
36 |
| [GGUF](https://huggingface.co/mradermacher/Swallow-MoE-2x13B-v0.1-GGUF/resolve/main/Swallow-MoE-2x13B-v0.1.Q2_K.gguf) | Q2_K | 8.1 | |
|
37 |
+
| [GGUF](https://huggingface.co/mradermacher/Swallow-MoE-2x13B-v0.1-GGUF/resolve/main/Swallow-MoE-2x13B-v0.1.Q3_K_S.gguf) | Q3_K_S | 9.5 | |
|
38 |
+
| [GGUF](https://huggingface.co/mradermacher/Swallow-MoE-2x13B-v0.1-GGUF/resolve/main/Swallow-MoE-2x13B-v0.1.Q3_K_M.gguf) | Q3_K_M | 10.6 | lower quality |
|
39 |
+
| [GGUF](https://huggingface.co/mradermacher/Swallow-MoE-2x13B-v0.1-GGUF/resolve/main/Swallow-MoE-2x13B-v0.1.Q3_K_L.gguf) | Q3_K_L | 11.5 | |
|
40 |
| [GGUF](https://huggingface.co/mradermacher/Swallow-MoE-2x13B-v0.1-GGUF/resolve/main/Swallow-MoE-2x13B-v0.1.Q4_K_S.gguf) | Q4_K_S | 12.4 | fast, recommended |
|
41 |
+
| [GGUF](https://huggingface.co/mradermacher/Swallow-MoE-2x13B-v0.1-GGUF/resolve/main/Swallow-MoE-2x13B-v0.1.Q4_K_M.gguf) | Q4_K_M | 13.2 | fast, recommended |
|
42 |
+
| [GGUF](https://huggingface.co/mradermacher/Swallow-MoE-2x13B-v0.1-GGUF/resolve/main/Swallow-MoE-2x13B-v0.1.Q6_K.gguf) | Q6_K | 17.8 | very good quality |
|
43 |
+
| [GGUF](https://huggingface.co/mradermacher/Swallow-MoE-2x13B-v0.1-GGUF/resolve/main/Swallow-MoE-2x13B-v0.1.Q8_0.gguf) | Q8_0 | 23.1 | fast, best quality |
|
44 |
|
45 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
46 |
types (lower is better):
|