Gemma-2-Llama-Swallow-9b-it-v0.1
tokyotech-llm/Gemma-2-Llama-Swallow-9b-it-v0.1
9BGemma 24 Heretic artifacts1 measured conversions18.6K downloads
Best version87.4Pareto
ChiKoi7/Gemma-2-Llama-Swallow-9b-it-v0.1-Hereticby ChiKoi7 · Heretic v1.1.0 · GGUF, SAFETENSORS
- KL divergence
- 0.238
- Refusals
- 6/100
- Refusal rate
- 6%
- Downloads
- 18.6K
Evidence from the model card
| Refusals (ja) | 6/100 | | KL divergence (ja) | 0.2380 |
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 | 9B | 0.238 | 6/1006% | 87.4 | BestParetoGGUFSAFETENSORS | 18.6K |
All artifacts (4)
| Repository | Type | Formats | Metrics | Downloads | Updated |
|---|---|---|---|---|---|
| Gemma-2-Llama-Swallow-9b-it-v0.1-Heretic-i1-GGUF mradermacher | quant | GGUFSAFETENSORSIQ1_M | inheritedKL 0.238 | 9.9K | 2025-12-18 |
| Gemma-2-Llama-Swallow-9b-it-v0.1-Heretic-GGUF ChiKoi7 | heretic | GGUF | published (weak match)KL 0.238 | 4.5K | 2025-12-16 |
| Gemma-2-Llama-Swallow-9b-it-v0.1-Heretic-GGUF mradermacher | quant | GGUFSAFETENSORSIQ4_XS | inheritedKL 0.238 | 3.8K | 2025-12-17 |
| Gemma-2-Llama-Swallow-9b-it-v0.1-Heretic ChiKoi7 | heretic | SAFETENSORS | published (weak match)KL 0.238 | 403 | 2025-12-16 |