Dataset Viewer
Auto-converted to Parquet Duplicate
model
string
suite
string
task
string
metric
string
bf16
float64
awq
float64
recovery_pct
float64
n
int64
ppl_ratio_awq_over_bf16
float64
value
float64
LostGentoo/Qwen3.5-4B-AWQ
openllm_lite
mmlu
acc
0.71
0.7
98.591549
200
null
null
LostGentoo/Qwen3.5-4B-AWQ
openllm_lite
arc_challenge
acc
0.545
0.555
101.834862
200
null
null
LostGentoo/Qwen3.5-4B-AWQ
openllm_lite
hellaswag
acc
0.705
0.67
95.035461
200
null
null
LostGentoo/Qwen3.5-4B-AWQ
openllm_lite
winogrande
acc
0.685
0.675
98.540146
200
null
null
LostGentoo/Qwen3.5-4B-AWQ
openllm_lite
truthfulqa_mc1
acc
0.295
0.285
96.610169
200
null
null
LostGentoo/Qwen3.5-4B-AWQ
openllm_lite
gsm8k
acc
0.02
0
0
50
null
null
LostGentoo/Qwen3.5-4B-AWQ
openllm_lite
wikitext
ppl
16.77
18.542
null
32
1.105665
null
LostGentoo/Qwen3-Embedding-8B-AWQ
mteb_lite
stsbenchmark
spearman
0.8844
0.881107
99.627709
null
null
null
LostGentoo/Qwen3-Embedding-8B-AWQ
mteb_lite
scifact
ndcg@10
0.7865
0.776822
98.769492
300
null
null
LostGentoo/Qwen3-Embedding-8B-AWQ
mteb_lite
sts_score_alignment
pearson_awq_vs_bf16
null
null
null
null
null
0.996448

AWQ Quant Quality Evals

Side-by-side AWQ W4A16 vs BF16 quality measurements for:

Quant Base Suite
LostGentoo/Qwen3.5-4B-AWQ Qwen/Qwen3.5-4B OpenLLM-lite
LostGentoo/Qwen3-Embedding-8B-AWQ Qwen/Qwen3-Embedding-8B MTEB-lite

Hardware: NVIDIA RTX 5060 Ti (sm_120, Blackwell).

Files

File Contents
quant_quality_evals.json Full combined report
qwen35_4b_awq_vs_bf16.json LLM OpenLLM-lite only
qwen3_embedding_8b_awq_vs_bf16.json Embedding MTEB-lite only
summary.jsonl Flat per-task rows (easy to load)

Headline results

Qwen3.5-4B (OpenLLM-lite, n=200 MC / task)

Task BF16 AWQ Recovery
MMLU 0.710 0.700 98.6%
ARC-Challenge 0.545 0.555 101.8%
HellaSwag 0.705 0.670 95.0%
Winogrande 0.685 0.675 98.5%
TruthfulQA MC1 0.295 0.285 96.6%
WikiText-2 PPL 16.77 18.54 +10.6%

Mean MC recovery: 98.1%.

GSM8K in this dump is not reliable (short gen budget + weak numeric extraction); treat as a harness caveat, not a quant cliff.

Qwen3-Embedding-8B (MTEB-lite)

Task BF16 AWQ Recovery
STSBenchmark Spearman 0.884 0.881 99.6%
SciFact nDCG@10 0.787 0.777 98.8%
STS score Pearson AWQ vs BF16 - - 0.996

Methodology

  • LLM MC tasks: greedy average-logprob scoring of answer continuations (ARC, HellaSwag, Winogrande, MMLU, TruthfulQA MC1), n=200 shuffled with seed 42.
  • WikiText-2 PPL: manual NLL from logits, 32 non-empty docs, max length 512.
  • Embeddings: last-token pooling + L2 normalize; retrieval queries use Qwen instruct prefix; SciFact nDCG@10 over test qrels.
  • Harness: custom script on top of cached HF datasets (lm-eval / mteb package install was unavailable in the run environment).

Load

from datasets import load_dataset
import json
from huggingface_hub import hf_hub_download

rows = load_dataset("LostGentoo/awq-quant-evals", split="train")
print(rows[0])

path = hf_hub_download("LostGentoo/awq-quant-evals", "quant_quality_evals.json", repo_type="dataset")
report = json.load(open(path))
print(report["qwen35_4b"]["recovery_pct"])

License

Apache-2.0. Eval numbers are measurements of third-party base models under their own licenses.

Downloads last month
44