Update README.md
Browse filesAdded VRAM requirement
README.md
CHANGED
|
@@ -1,3 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
base_model:
|
| 3 |
- openbmb/Eurux-8x22b-nca
|
|
|
|
| 1 |
+
3.75bpw EXL2 quant of https://huggingface.co/gghfez/WizardLM-2-8x22B-Beige
|
| 2 |
+
|
| 3 |
+
This is the biggest quant we can fit into 72GB of VRAM (eg. 3x3090 cards) with a Q4 cache
|
| 4 |
+
|
| 5 |
---
|
| 6 |
base_model:
|
| 7 |
- openbmb/Eurux-8x22b-nca
|