Qwen3-4B-Thinking-2507
unsloth/Qwen3-4B-Thinking-2507
4BQwen 37 Heretic artifacts1 measured conversions555 downloads
Best version97.9Pareto
JoaoZaokk/Qwen3-4B-Thinking-2507-MiniMax-M2.1-Distill-hereticby JoaoZaokk · Heretic v1.3.0 · AWQ, GGUF, GPTQ, LORA, SAFETENSORS
- KL divergence
- 0.0025
- Refusals
- 4/100
- Refusal rate
- 4%
- Downloads
- 555
Evidence from the model card
| KL divergence | 0.0025 | | Refusals | 4/100 |
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 | JoaoZaokk | 4B | 0.0025 | 4/1004% | 97.9 | BestParetoAWQGGUFGPTQ | 555 |
All artifacts (7)
| Repository | Type | Formats | Metrics | Downloads | Updated |
|---|---|---|---|---|---|
| Qwen3-4B-Thinking-2507-Heretic-CodeFeedback JoaoZaokk | lora | LORASAFETENSORS | inheritedKL 0.0025 | 170 | 2026-06-06 |
| Qwen3-4B-Thinking-2507-Heretic-CodeFeedback-OpenVINO-INT4 JoaoZaokk | merge | OTHERINT4 | inheritedKL 0.0025 | 128 | 2026-06-06 |
| Qwen3-4B-Thinking-2507-Heretic-CodeFeedback-Agentic-Tessa-1K-LoRA JoaoZaokk | lora | LORASAFETENSORS | inheritedKL 0.0025 | 69 | 2026-06-09 |
| Qwen3-4B-Thinking-2507-MiniMax-M2.1-Distill-heretic JoaoZaokk | heretic | SAFETENSORS | published (weak match)KL 0.0025 | 68 | 2026-06-04 |
| Qwen3-4B-Thinking-2507-Heretic-CodeFeedback-OpenCodeInstruct-Learning-LoRA JoaoZaokk | lora | LORASAFETENSORS | inheritedKL 0.0025 | 48 | 2026-06-07 |
| Qwen3-4B-Thinking-2507-Heretic-CodeFeedback-Agentic-Tessa-8K-2048-Experimental-LoRA JoaoZaokk | lora | LORASAFETENSORS | inheritedKL 0.0025 | 43 | 2026-06-09 |
| Qwen3-4B-Thinking-2507-Heretic-CodeFeedback-OpenVINO-INT8 JoaoZaokk | finetune | OTHERINT8 | inheritedKL 0.0025 | 29 | 2026-09-04 |