Downloads · 30 days
18
12% of all-time downloads
Academica/Academica-Qwen3-8B-2Bit-Heretic
Academica-Qwen3-8B-2Bit-Heretic is a text generation model from Academica. Use it when you need the model to write or continue text. The card lists the license as apache-2.0.
Qwen3-8B with Heretic refusal-direction ablation and importance-weighted IQ1S quantization.
Downloads · 30 days
18
12% of all-time downloads
All-time downloads
149
Public
Repo size
2.1 GB
Likes
0
Public
Click a slice to open those files.
.gguf2.1 GB · 100%
From the Hugging Face model README
Qwen3-8B with Heretic refusal-direction ablation and importance-weighted IQ1_S quantization.
Academica produced the ablated model and all three published quantizations. This 2-bit release prioritizes minimum memory use and maximum throughput.
| Property | Value |
|---|---|
| Base model | Qwen/Qwen3-8B |
| Modification | Heretic ablation after 20 Optuna trials |
| Quantization | IQ1_S, importance weighted |
| File size | 2.0 GB |
| Measured throughput | 95 tokens/second |
| Measured perplexity | 2.06 |
| Refusal evaluation | 18 refusals from 100 prompts |
| Evaluation hardware | NVIDIA DGX Spark, GB10 Blackwell, 128 GB unified memory |
The measurements describe the published artifact on the stated hardware. Results can differ with prompts, runtimes, and sampling settings.
qwen3-8b-heretic-iq1_s.gguf: IQ1_S model weights.qwen3-8b-heretic-imatrix-v2.dat: importance matrix used during quantization.from huggingface_hub import hf_hub_download
model_path = hf_hub_download(
repo_id="Academica/Academica-Qwen3-8B-2Bit-Heretic",
filename="qwen3-8b-heretic-iq1_s.gguf",
)
llama-server -m qwen3-8b-heretic-iq1_s.gguf --host 0.0.0.0 --port 8080 -ngl 99
IQ1_S compression can reduce coherence and increase repetition. The 3-bit release provides a stronger quality and size balance.
The ablation reduces refusal behavior. It does not establish factual accuracy, safety, or suitability for regulated decisions.