DeepHermes-3-Llama-3-8B-Preview
NousResearch/DeepHermes-3-Llama-3-8B-Preview
8BLlama 33 Heretic artifacts1 measured conversions20K downloads
Best version96.9Pareto
ChiKoi7/DeepHermes-3-Llama-3-8B-Preview-Hereticby ChiKoi7 · Heretic v1.1.0 · GGUF, SAFETENSORS
- KL divergence
- 0.023
- Refusals
- 4/100
- Refusal rate
- 4%
- Downloads
- 20K
Evidence from the model card
| Refusals | 4/100 | | KL divergence | 0.0226 |
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 | ChiKoi7 | 8B | 0.023 | 4/1004% | 96.9 | BestParetoGGUFSAFETENSORS | 20K |
All artifacts (3)
| Repository | Type | Formats | Metrics | Downloads | Updated |
|---|---|---|---|---|---|
| DeepHermes-3-Llama-3-8B-Preview-Heretic-i1-GGUF mradermacher | quant | GGUFIQ1_M | inheritedKL 0.023 | 15.4K | 2025-12-24 |
| DeepHermes-3-Llama-3-8B-Preview-Heretic-GGUF mradermacher | quant | GGUFIQ4_XS | inheritedKL 0.023 | 4.5K | 2025-12-24 |
| DeepHermes-3-Llama-3-8B-Preview-Heretic ChiKoi7 | heretic | SAFETENSORS | published (weak match)KL 0.023 | 97 | 2025-12-24 |