SpiridonSunRotator commited on
Commit
509ed2d
1 Parent(s): cdeead5

Added fp16 baseline

Browse files
Files changed (1) hide show
  1. README.md +4 -3
README.md CHANGED
@@ -3,9 +3,10 @@ Official [AQLM](https://arxiv.org/abs/2401.06118) quantization of [google/gemma-
3
  For this quantization, we used 1 codebook of 16 bits.
4
 
5
 
6
- Results:
7
- | Model | AQLM scheme | WinoGrande | PiQA | HellaSwag | ArcE | ArcC | Model size, Gb |
8
  |------|------|------|-------|-------|-------|------|------|
9
- | gemma-2b |1x16| 0.6275 | 0.7318 | 0.4582 | 0.6923 | 0.3259| 1.7 |
 
10
 
11
  To learn more about the inference, as well as the information on how to quantize models yourself, please refer to the [official GitHub repo](https://github.com/Vahe1994/AQLM).
 
3
  For this quantization, we used 1 codebook of 16 bits.
4
 
5
 
6
+ Results (0-shot `acc`):
7
+ | Model | Quantization | WinoGrande | PiQA | HellaSwag | ArcE | ArcC | Model size, Gb |
8
  |------|------|------|-------|-------|-------|------|------|
9
+ |gemma-2b| None | 0.6472 | 0.7715 | 0.5279 | 0.7403 | 0.4053| 5.0 |
10
+ | |1x16| 0.6275 | 0.7318 | 0.4582 | 0.6923 | 0.3259| 1.7 |
11
 
12
  To learn more about the inference, as well as the information on how to quantize models yourself, please refer to the [official GitHub repo](https://github.com/Vahe1994/AQLM).