Commit
•
230b5b0
1
Parent(s):
e1a47fb
auto-patch README.md
Browse files
README.md
CHANGED
@@ -37,7 +37,7 @@ quantized_by: mradermacher
|
|
37 |
static quants of https://huggingface.co/CohereForAI/aya-23-8B
|
38 |
|
39 |
<!-- provided-files -->
|
40 |
-
weighted/imatrix quants
|
41 |
## Usage
|
42 |
|
43 |
If you are unsure how to use GGUF files, refer to one of [TheBloke's
|
@@ -50,7 +50,21 @@ more details, including on how to concatenate multi-part files.
|
|
50 |
|
51 |
| Link | Type | Size/GB | Notes |
|
52 |
|:-----|:-----|--------:|:------|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
53 |
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.Q8_0.gguf) | Q8_0 | 8.6 | fast, best quality |
|
|
|
54 |
|
55 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
56 |
types (lower is better):
|
|
|
37 |
static quants of https://huggingface.co/CohereForAI/aya-23-8B
|
38 |
|
39 |
<!-- provided-files -->
|
40 |
+
weighted/imatrix quants are available at https://huggingface.co/mradermacher/aya-23-8B-i1-GGUF
|
41 |
## Usage
|
42 |
|
43 |
If you are unsure how to use GGUF files, refer to one of [TheBloke's
|
|
|
50 |
|
51 |
| Link | Type | Size/GB | Notes |
|
52 |
|:-----|:-----|--------:|:------|
|
53 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.Q2_K.gguf) | Q2_K | 3.5 | |
|
54 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.IQ3_XS.gguf) | IQ3_XS | 3.8 | |
|
55 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.Q3_K_S.gguf) | Q3_K_S | 4.0 | |
|
56 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.IQ3_S.gguf) | IQ3_S | 4.0 | beats Q3_K* |
|
57 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.IQ3_M.gguf) | IQ3_M | 4.1 | |
|
58 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.Q3_K_M.gguf) | Q3_K_M | 4.3 | lower quality |
|
59 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.Q3_K_L.gguf) | Q3_K_L | 4.6 | |
|
60 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.IQ4_XS.gguf) | IQ4_XS | 4.7 | |
|
61 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.Q4_K_S.gguf) | Q4_K_S | 4.9 | fast, recommended |
|
62 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.Q4_K_M.gguf) | Q4_K_M | 5.2 | fast, recommended |
|
63 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.Q5_K_S.gguf) | Q5_K_S | 5.8 | |
|
64 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.Q5_K_M.gguf) | Q5_K_M | 5.9 | |
|
65 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.Q6_K.gguf) | Q6_K | 6.7 | very good quality |
|
66 |
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.Q8_0.gguf) | Q8_0 | 8.6 | fast, best quality |
|
67 |
+
| [GGUF](https://huggingface.co/mradermacher/aya-23-8B-GGUF/resolve/main/aya-23-8B.f16.gguf) | f16 | 16.2 | 16 bpw, overkill |
|
68 |
|
69 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
70 |
types (lower is better):
|