Llama-SEA-LION-v3-8B-IT
aisingapore/Llama-SEA-LION-v3-8B-IT
8BLlama3 Heretic artifacts1 measured conversions16.3K downloads
Best version94.0Pareto
MagicalAlchemist/Llama-SEA-LION-v3-8B-IT-Magic_decensoredby MagicalAlchemist · Heretic v1.1.0 · GGUF, SAFETENSORS
- KL divergence
- 0.031
- Refusals
- 9/100
- Refusal rate
- 9%
- Downloads
- 16.3K
Evidence from the model card
| KL divergence | 0.0308 | | Refusals | 9/100 |
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 | MagicalAlchemist | 8B | 0.031 | 9/1009% | 94.0 | BestParetoGGUFSAFETENSORS | 16.3K |
All artifacts (3)
| Repository | Type | Formats | Metrics | Downloads | Updated |
|---|---|---|---|---|---|
| Llama-SEA-LION-v3-8B-IT-Magic_decensored-i1-GGUF mradermacher | quant | GGUFIQ1_M | inheritedKL 0.031 | 14.3K | 2026-01-30 |
| Llama-SEA-LION-v3-8B-IT-Magic_decensored-GGUF mradermacher | quant | GGUFIQ4_XS | inheritedKL 0.031 | 1.6K | 2026-01-30 |
| Llama-SEA-LION-v3-8B-IT-Magic_decensored MagicalAlchemist | heretic | SAFETENSORS | published (weak match)KL 0.031 | 408 | 2026-01-25 |