LFM2.5-1.2B-Instruct-heretic-GGUF
GGUF quantizations of richardyoung/LFM2.5-1.2B-Instruct-heretic, a Heretic abliteration of LiquidAI/LFM2.5-1.2B-Instruct.
| Metric (Heretic evaluation) | Value |
|---|---|
| KL divergence from the original | 0.0585 |
| Refusals | 3/100 |
Reproducibility information: richardyoung/LFM2.5-1.2B-Instruct-heretic/reproduce.
| File | Quant |
|---|---|
LFM2.5-1.2B-Instruct-heretic-Q4_K_M.gguf |
Q4_K_M |
LFM2.5-1.2B-Instruct-heretic-Q5_K_M.gguf |
Q5_K_M |
LFM2.5-1.2B-Instruct-heretic-Q6_K.gguf |
Q6_K |
LFM2.5-1.2B-Instruct-heretic-Q8_0.gguf |
Q8_0 |
Ollama: ollama run richardyoung/lfm2.5-1.2b-instruct-heretic (default tag Q4_K_M).