Update README.md
Browse files
README.md
CHANGED
@@ -27,12 +27,12 @@ A Convex frankenMoE. Created via improving the original Seraphim script. The mod
|
|
27 |
|
28 |
| Name | Quant method | Bits | Size | Max RAM required | Use case |
|
29 |
| ---- | ---- | ---- | ---- | ---- | ----- |
|
30 |
-
| [Q2_K Tiny](https://huggingface.co/Kquant03/
|
31 |
-
| [Q3_K_M](https://huggingface.co/Kquant03/
|
32 |
-
| [Q4_0](https://huggingface.co/Kquant03/
|
33 |
-
| [Q4_K_M](https://huggingface.co/Kquant03/
|
34 |
-
| [Q5_0](https://huggingface.co/Kquant03/
|
35 |
-
| [Q5_K_M](https://huggingface.co/Kquant03/
|
36 |
# "[What is a Mixture of Experts (MoE)?](https://huggingface.co/blog/moe)"
|
37 |
### (from the MistralAI papers...click the quoted question above to navigate to it directly.)
|
38 |
|
|
|
27 |
|
28 |
| Name | Quant method | Bits | Size | Max RAM required | Use case |
|
29 |
| ---- | ---- | ---- | ---- | ---- | ----- |
|
30 |
+
| [Q2_K Tiny](https://huggingface.co/Kquant03/BurningBruce-SOLAR-8x10.7B-GGUF/blob/main/ggml-model-q2_k.gguf) | Q2_K | 2 | 23.4 GB| 25.4 GB | smallest, significant quality loss - not recommended for most purposes |
|
31 |
+
| [Q3_K_M](https://huggingface.co/Kquant03/BurningBruce-SOLAR-8x10.7B-GGUF/blob/main/ggml-model-q3_k_m.gguf) | Q3_K_M | 3 | 30.5 GB| 32.5 GB | very small, high quality loss |
|
32 |
+
| [Q4_0](https://huggingface.co/Kquant03/BurningBruce-SOLAR-8x10.7B-GGUF/blob/main/ggml-model-q4_0.gguf) | Q4_0 | 4 | 39.6 GB| 41.6 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
|
33 |
+
| [Q4_K_M](https://huggingface.co/Kquant03/BurningBruce-SOLAR-8x10.7B-GGUF/blob/main/ggml-model-q4_k_m.gguf) | Q4_K_M | 4 | ~39.6 GB| ~41.6 GB | medium, balanced quality - recommended |
|
34 |
+
| [Q5_0](https://huggingface.co/Kquant03/BurningBruce-SOLAR-8x10.7B-GGUF/blob/main/ggml-model-q5_0.gguf) | Q5_0 | 5 | 48.2 GB| 50.2 GB | legacy; large, balanced quality |
|
35 |
+
| [Q5_K_M](https://huggingface.co/Kquant03/BurningBruce-SOLAR-8x10.7B-GGUF/blob/main/ggml-model-q5_k_m.gguf) | Q5_K_M | 5 | ~48.2 GB| ~50.2 GB | large, balanced quality - recommended |
|
36 |
# "[What is a Mixture of Experts (MoE)?](https://huggingface.co/blog/moe)"
|
37 |
### (from the MistralAI papers...click the quoted question above to navigate to it directly.)
|
38 |
|