Commit
•
4450093
1
Parent(s):
be5a887
auto-patch README.md
Browse files
README.md
CHANGED
|
@@ -38,16 +38,26 @@ more details, including on how to concatenate multi-part files.
|
|
| 38 |
|
| 39 |
| Link | Type | Size/GB | Notes |
|
| 40 |
|:-----|:-----|--------:|:------|
|
|
|
|
| 41 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ1_M.gguf) | i1-IQ1_M | 13.6 | mostly desperate |
|
|
|
|
|
|
|
|
|
|
| 42 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ2_M.gguf) | i1-IQ2_M | 19.3 | |
|
| 43 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q2_K.gguf) | i1-Q2_K | 20.8 | IQ3_XXS probably better |
|
| 44 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 21.2 | lower quality |
|
|
|
|
|
|
|
| 45 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q3_K_S.gguf) | i1-Q3_K_S | 24.2 | IQ3_XS probably better |
|
| 46 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ3_M.gguf) | i1-IQ3_M | 25.4 | |
|
| 47 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q3_K_M.gguf) | i1-Q3_K_M | 26.9 | IQ3_S probably better |
|
|
|
|
| 48 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ4_XS.gguf) | i1-IQ4_XS | 29.6 | |
|
|
|
|
| 49 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q4_K_S.gguf) | i1-Q4_K_S | 31.5 | optimal size/speed/quality |
|
| 50 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q4_K_M.gguf) | i1-Q4_K_M | 33.4 | fast, recommended |
|
|
|
|
|
|
|
| 51 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q6_K.gguf) | i1-Q6_K | 45.1 | practically like static Q6_K |
|
| 52 |
|
| 53 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
|
|
|
| 38 |
|
| 39 |
| Link | Type | Size/GB | Notes |
|
| 40 |
|:-----|:-----|--------:|:------|
|
| 41 |
+
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ1_S.gguf) | i1-IQ1_S | 12.6 | for the desperate |
|
| 42 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ1_M.gguf) | i1-IQ1_M | 13.6 | mostly desperate |
|
| 43 |
+
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 15.3 | |
|
| 44 |
+
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ2_XS.gguf) | i1-IQ2_XS | 16.7 | |
|
| 45 |
+
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ2_S.gguf) | i1-IQ2_S | 18.0 | |
|
| 46 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ2_M.gguf) | i1-IQ2_M | 19.3 | |
|
| 47 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q2_K.gguf) | i1-Q2_K | 20.8 | IQ3_XXS probably better |
|
| 48 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 21.2 | lower quality |
|
| 49 |
+
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ3_XS.gguf) | i1-IQ3_XS | 22.9 | |
|
| 50 |
+
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ3_S.gguf) | i1-IQ3_S | 24.2 | beats Q3_K* |
|
| 51 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q3_K_S.gguf) | i1-Q3_K_S | 24.2 | IQ3_XS probably better |
|
| 52 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ3_M.gguf) | i1-IQ3_M | 25.4 | |
|
| 53 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q3_K_M.gguf) | i1-Q3_K_M | 26.9 | IQ3_S probably better |
|
| 54 |
+
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q3_K_L.gguf) | i1-Q3_K_L | 29.4 | IQ3_M probably better |
|
| 55 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-IQ4_XS.gguf) | i1-IQ4_XS | 29.6 | |
|
| 56 |
+
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q4_0.gguf) | i1-Q4_0 | 31.4 | fast, low quality |
|
| 57 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q4_K_S.gguf) | i1-Q4_K_S | 31.5 | optimal size/speed/quality |
|
| 58 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q4_K_M.gguf) | i1-Q4_K_M | 33.4 | fast, recommended |
|
| 59 |
+
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q5_K_S.gguf) | i1-Q5_K_S | 38.0 | |
|
| 60 |
+
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q5_K_M.gguf) | i1-Q5_K_M | 39.1 | |
|
| 61 |
| [GGUF](https://huggingface.co/mradermacher/Qwen1.5-55B-Chat-Cut-i1-GGUF/resolve/main/Qwen1.5-55B-Chat-Cut.i1-Q6_K.gguf) | i1-Q6_K | 45.1 | practically like static Q6_K |
|
| 62 |
|
| 63 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|