Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-nc-4.0
|
| 3 |
+
---
|
| 4 |
+
## Description
|
| 5 |
+
|
| 6 |
+
Exllama 2 quant of [Undi95/Utopia-13B](https://huggingface.co/Undi95/Utopia-13B)
|
| 7 |
+
|
| 8 |
+
8 BPW, Head bit set to 8
|
| 9 |
+
|
| 10 |
+
## Prompt template: Alpaca
|
| 11 |
+
|
| 12 |
+
```
|
| 13 |
+
Below is an instruction that describes a task. Write a response that appropriately completes the request.
|
| 14 |
+
|
| 15 |
+
### Instruction:
|
| 16 |
+
{prompt}
|
| 17 |
+
|
| 18 |
+
### Response:
|
| 19 |
+
|
| 20 |
+
```
|
| 21 |
+
|
| 22 |
+
## VRAM
|
| 23 |
+
My VRAM usage with 13B models are:
|
| 24 |
+
| Bits per weight | Context | VRAM |
|
| 25 |
+
|--|--|--|
|
| 26 |
+
| 8bpw | 8k | 22gb |
|
| 27 |
+
| 8bpw | 4k | 19gb |
|
| 28 |
+
| 6bpw | 8k | 19gb |
|
| 29 |
+
| 6bpw | 4k | 16gb |
|
| 30 |
+
| 4bpw | 8k | 16gb |
|
| 31 |
+
| 4bpw | 4k | 13gb |
|
| 32 |
+
| 3bpw | 8k | 15gb |
|
| 33 |
+
| 3bpw | 4k | 12gb |
|
| 34 |
+
I have rounded up, these arent exact numbers, this is also on a windows machine, they should be slightly lower on linux.
|