Llama-3.1-Nemotron-Nano-4B-v1.1
nvidia/Llama-3.1-Nemotron-Nano-4B-v1.1
4BLlama 3.13 Heretic artifacts1 measured conversions6.7K downloads
Best version95.7Pareto
hardcoreerik/Llama-3.1-Nemotron-Nano-4B-v1.1-hereticby hardcoreerik · Heretic v1.3.0 · GGUF, SAFETENSORS
- KL divergence
- 0.016
- Refusals
- 7/100
- Refusal rate
- 7%
- Downloads
- 6.7K
Evidence from the model card
| KL divergence | 0.0157 | | Refusals | 7/100 |
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 | hardcoreerik | 4B | 0.016 | 7/1007% | 95.7 | BestParetoGGUFSAFETENSORS | 6.7K |
All artifacts (3)
| Repository | Type | Formats | Metrics | Downloads | Updated |
|---|---|---|---|---|---|
| Llama-3.1-Nemotron-Nano-4B-v1.1-heretic-i1-GGUF mradermacher | quant | GGUFIQ1_M | inheritedKL 0.016 | 4.7K | 2026-06-08 |
| Llama-3.1-Nemotron-Nano-4B-v1.1-heretic-GGUF mradermacher | quant | GGUFIQ4_XS | inheritedKL 0.016 | 1.7K | 2026-06-08 |
| Llama-3.1-Nemotron-Nano-4B-v1.1-heretic hardcoreerik | heretic | SAFETENSORS | published (weak match)KL 0.016 | 332 | 2026-06-08 |