Qwen2.5-VL-3B-Instruct
Qwen/Qwen2.5-VL-3B-Instruct
3BQwen 2.56 Heretic artifacts4 measured conversions9.5K downloads
Best version98.1Pareto
saidutta69/Qwen2.5-VL-3B-Instruct-hereticby saidutta69 · Heretic v1.4.0 · GGUF, SAFETENSORS
- KL divergence
- 0.018
- Refusals
- 2/100
- Refusal rate
- 2%
- Downloads
- 726
Evidence from the model card
| KL divergence | 0.0178 | | Refusals | 2/100 |
Metrics published by the author. Why this ranks first.
All measured versions
Higher score is better. One row per ablation attempt; artifacts of the same attempt are grouped under it.
| # | Author | Params | KL | Refusals | Score | Evidence | Downloads |
|---|---|---|---|---|---|---|---|
| 1 | saidutta69 | 3B | 0.018 | 2/1002% | 98.1 | BestParetoGGUFSAFETENSORS | 726 |
| 2 | sh0ck0r | 3B | 0.022 | 3/1003% | 97.4 | SAFETENSORS | 104 |
| 3 | coder3101 | 3B | 0.010 | 5/1005% | 97.0 | ParetoGGUFSAFETENSORS | 8.6K |
| 4 | genevera | 3B | 0.090 | 4/1004% | 93.9 | SAFETENSORS | 128 |
All artifacts (6)
| Repository | Type | Formats | Metrics | Downloads | Updated |
|---|---|---|---|---|---|
| Qwen2.5-VL-3B-Instruct-heretic-i1-GGUF mradermacher | quant | GGUFIQ1_M | inheritedKL 0.010 | 4.1K | 2026-01-18 |
| Qwen2.5-VL-3B-Instruct-heretic-GGUF mradermacher | quant | GGUFIQ4_XS | inheritedKL 0.010 | 3.9K | 2026-10-08 |
| Qwen2.5-VL-3B-Instruct-heretic saidutta69 | heretic | GGUFSAFETENSORSF16 | published (weak match)KL 0.018 | 726 | 2026-10-13 |
| Qwen2.5-VL-3B-Instruct-heretic coder3101 | heretic | SAFETENSORS | published (weak match)KL 0.010 | 599 | 2026-01-17 |
| Qwen2.5-VL-3B-Instruct-heretic genevera | heretic | SAFETENSORS | published (weak match)KL 0.090 | 128 | 2025-12-07 |
| Qwen2.5-VL-3B-Instruct-heretic sh0ck0r | heretic | SAFETENSORS | published (weak match)KL 0.022 | 104 | 2026-03-07 |