mradermacher
commited on
Commit
•
a6dcb55
1
Parent(s):
ce26743
auto-patch README.md
Browse files
README.md
CHANGED
@@ -42,7 +42,21 @@ more details, including on how to concatenate multi-part files.
|
|
42 |
|
43 |
| Link | Type | Size/GB | Notes |
|
44 |
|:-----|:-----|--------:|:------|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
45 |
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.Q8_0.gguf) | Q8_0 | 1.7 | fast, best quality |
|
|
|
46 |
|
47 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
48 |
types (lower is better):
|
|
|
42 |
|
43 |
| Link | Type | Size/GB | Notes |
|
44 |
|:-----|:-----|--------:|:------|
|
45 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.Q2_K.gguf) | Q2_K | 0.8 | |
|
46 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.IQ3_XS.gguf) | IQ3_XS | 0.8 | |
|
47 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.Q3_K_S.gguf) | Q3_K_S | 0.9 | |
|
48 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.IQ3_S.gguf) | IQ3_S | 0.9 | beats Q3_K* |
|
49 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.IQ3_M.gguf) | IQ3_M | 0.9 | |
|
50 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.Q3_K_M.gguf) | Q3_K_M | 0.9 | lower quality |
|
51 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.Q3_K_L.gguf) | Q3_K_L | 1.0 | |
|
52 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.IQ4_XS.gguf) | IQ4_XS | 1.0 | |
|
53 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.Q4_K_S.gguf) | Q4_K_S | 1.0 | fast, recommended |
|
54 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.Q4_K_M.gguf) | Q4_K_M | 1.1 | fast, recommended |
|
55 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.Q5_K_S.gguf) | Q5_K_S | 1.2 | |
|
56 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.Q5_K_M.gguf) | Q5_K_M | 1.2 | |
|
57 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.Q6_K.gguf) | Q6_K | 1.4 | very good quality |
|
58 |
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.Q8_0.gguf) | Q8_0 | 1.7 | fast, best quality |
|
59 |
+
| [GGUF](https://huggingface.co/mradermacher/SeaLLMs-v3-1.5B-GGUF/resolve/main/SeaLLMs-v3-1.5B.f16.gguf) | f16 | 3.2 | 16 bpw, overkill |
|
60 |
|
61 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
62 |
types (lower is better):
|