Downloads · 30 days
137
24% of all-time downloads
Qwen3.8-27B-NVFP4-FP8-GPTQ-AWQ
dsikka/Qwen3.8-27B-NVFP4-FP8-GPTQ-AWQ
Qwen3.8-27B-NVFP4-FP8-GPTQ-AWQ is a machine learning model from dsikka. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
All-time downloads
581
Public
Parameters
19.9B
22.6 GB on disk
Likes
1
Public
Repo makeup
Click a slice to open those files.
.safetensors22.6 GB · 100%
Parameter types
How the weights are stored.
F8_E4M310.6B · 53%
Model card
From the Hugging Face model README
| Benchmark (metric) | Inferact | Unsloth | This checkpoint |
|---|---|---|---|
| gpqa_diamond (gpqa_pass@k:k=1) | 88.05 | 88.05 | 89.39 |
| aime25 (pass@k:k=1&n=1) | 90.00 | 93.75 | 97.08 |
| gsm8k_platinum (exact_match,strict-match) | 92.69 | 95.51 | 96.66 |
| mmlu_pro_chat (exact_match,custom-extract) | 82.77 | 83.88 | 83.95 |
| ifeval (prompt_level_strict_acc) | 91.37 | 91.74 | 91.50 |
| math_500 (pass@k:k=1&n=1) | 83.13 | 85.33 | 84.07 |
- Type
- qwen3_5
- Created
- Aug 17, 2026
- Updated
- Aug 17, 2026