mradermacher
commited on
Commit
•
facd43a
1
Parent(s):
086bf6f
auto-patch README.md
Browse files
README.md
CHANGED
@@ -22,7 +22,7 @@ tags:
|
|
22 |
weighted/imatrix quants of https://huggingface.co/EpistemeAI/Athena-codegemma-2-2b-it
|
23 |
|
24 |
<!-- provided-files -->
|
25 |
-
static quants are available at https://huggingface.co/mradermacher/Athena-
|
26 |
## Usage
|
27 |
|
28 |
If you are unsure how to use GGUF files, refer to one of [TheBloke's
|
@@ -35,27 +35,27 @@ more details, including on how to concatenate multi-part files.
|
|
35 |
|
36 |
| Link | Type | Size/GB | Notes |
|
37 |
|:-----|:-----|--------:|:------|
|
38 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
39 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
40 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
41 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
42 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
43 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
44 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
45 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
46 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
47 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
48 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
49 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
50 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
51 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
52 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
53 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
54 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
55 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
56 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
57 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
58 |
-
| [GGUF](https://huggingface.co/mradermacher/Athena-
|
59 |
|
60 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
61 |
types (lower is better):
|
|
|
22 |
weighted/imatrix quants of https://huggingface.co/EpistemeAI/Athena-codegemma-2-2b-it
|
23 |
|
24 |
<!-- provided-files -->
|
25 |
+
static quants are available at https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-GGUF
|
26 |
## Usage
|
27 |
|
28 |
If you are unsure how to use GGUF files, refer to one of [TheBloke's
|
|
|
35 |
|
36 |
| Link | Type | Size/GB | Notes |
|
37 |
|:-----|:-----|--------:|:------|
|
38 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-IQ1_S.gguf) | i1-IQ1_S | 0.9 | for the desperate |
|
39 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-IQ1_M.gguf) | i1-IQ1_M | 1.0 | mostly desperate |
|
40 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 1.0 | |
|
41 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-IQ2_XS.gguf) | i1-IQ2_XS | 1.1 | |
|
42 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-IQ2_S.gguf) | i1-IQ2_S | 1.1 | |
|
43 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-IQ2_M.gguf) | i1-IQ2_M | 1.2 | |
|
44 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 1.3 | lower quality |
|
45 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-Q2_K.gguf) | i1-Q2_K | 1.3 | IQ3_XXS probably better |
|
46 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-IQ3_XS.gguf) | i1-IQ3_XS | 1.4 | |
|
47 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-IQ3_S.gguf) | i1-IQ3_S | 1.5 | beats Q3_K* |
|
48 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-Q3_K_S.gguf) | i1-Q3_K_S | 1.5 | IQ3_XS probably better |
|
49 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-IQ3_M.gguf) | i1-IQ3_M | 1.5 | |
|
50 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-Q3_K_M.gguf) | i1-Q3_K_M | 1.6 | IQ3_S probably better |
|
51 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-Q3_K_L.gguf) | i1-Q3_K_L | 1.7 | IQ3_M probably better |
|
52 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-IQ4_XS.gguf) | i1-IQ4_XS | 1.7 | |
|
53 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-Q4_0.gguf) | i1-Q4_0 | 1.7 | fast, low quality |
|
54 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-Q4_K_S.gguf) | i1-Q4_K_S | 1.7 | optimal size/speed/quality |
|
55 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-Q4_K_M.gguf) | i1-Q4_K_M | 1.8 | fast, recommended |
|
56 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-Q5_K_S.gguf) | i1-Q5_K_S | 2.0 | |
|
57 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-Q5_K_M.gguf) | i1-Q5_K_M | 2.0 | |
|
58 |
+
| [GGUF](https://huggingface.co/mradermacher/Athena-gemma-2-2b-it-i1-GGUF/resolve/main/Athena-codegemma-2-2b-it.i1-Q6_K.gguf) | i1-Q6_K | 2.3 | practically like static Q6_K |
|
59 |
|
60 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
61 |
types (lower is better):
|