Hugging Face's logo

ISTA-DASLab
/
Meta-Llama-3-8B-Instruct-AQLM-2Bit-1x16

Text Generation
Transformers
Safetensors
llama
facebook
meta
llama-3
conversational
text-generation-inference
Inference Endpoints
aqlm
Model card Files Community
4
SpiridonSunRotator commited on
Commit
90f515b
·
verified ·
1 Parent(s): 7d8c61e

Added more evaluations

Browse files
Files changed (1) hide show
  1. README.md +4 -4
README.md CHANGED
@@ -14,7 +14,7 @@ Official [AQLM](https://arxiv.org/abs/2401.06118) quantization of [meta-llama/Me
14
  For this quantization, we used 1 codebook of 16 bits.
15
 
16
  Results:
17
- | Model | Quantization | MMLU (5-shot) | GSM8k (8-shot) | Model size, Gb |
18
- |------|------|-------|------|------|
19
- |meta-llama/Meta-Llama-3-8B-Instruct| None |0.6560 | 0.7475 | 16.1 |
20
- | | 1x16 | 0.5529 | 0.3397 | 4.1 |
 
14
  For this quantization, we used 1 codebook of 16 bits.
15
 
16
  Results:
17
+ | Model | Quantization | MMLU (5-shot) | GSM8k (8-shot) | ArcC| ArcE| Hellaswag | Winogrande | PiQA | Model size, Gb |
18
+ |------|------|-------|------|------|------|------|------|------|------|
19
+ |meta-llama/Meta-Llama-3-8B-Instruct| None |0.6560 | 0.7475 | 0.5299 | 0.8165 | 0.5771 | 0.7867 | 0.7206 | 16.1 |
20
+ | | 1x16 | 0.5529 | 0.3397 | 0.4428 | 0.7580 | 0.5313 | 0.7666 | 0.6788 | 4.1 |