base_model: ornith-ai/Ornith-1.5-9B
license: mit
pipeline_tag: text-generation
tags:
- gguf
- llama.cpp
- reasoning
- uncensored
Ornith-1.5-9B-Uncensored-GGUF
GGUF quantizations of an uncensored variant of
ornith-ai/Ornith-1.5-9B.
These GGUF files contain the text model only. Vision projector and MTP weights are not included.
Files and evaluation
| File | Size | JailbreakBench ASR | HarmBench ASR (159 standard behaviors) |
|---|---|---|---|
Ornith-1.5-9B-Uncensored-Q8_0.gguf |
9.53 GB | 95/100 (95.0%) | 153/159 (96.2%) |
Ornith-1.5-9B-Uncensored-Q6_K.gguf |
7.36 GB | 94/100 (94.0%) | 149/159 (93.7%) |
Ornith-1.5-9B-Uncensored-Q5_K_M.gguf |
6.47 GB | 91/100 (91.0%) | 148/159 (93.1%) |
Ornith-1.5-9B-Uncensored-Q4_K_M.gguf |
5.63 GB | 90/100 (90.0%) | 146/159 (91.8%) |
Higher ASR indicates fewer refusals on these benchmarks. HarmBench behaviors were submitted as
direct requests. Evaluation used reasoning off, temperature 0, seed 42, a 256-token output limit,
and the HarmBench-Llama-2-13b-cls classifier. Results can vary with runtime, prompt format,
sampling, and context.
Usage
llama-server \
--model /path/to/model.gguf \
--host 0.0.0.0 \
--port 8080 \
--ctx-size 16384 \
--reasoning on
The model supports a maximum context length of 262,144 tokens. Usable context depends on
available memory and runtime settings.
License
MIT. See LICENSE, NOTICE, and the
original model for upstream information.