ussoewwin/qwen3_4b_8b_abliterated_fp16

Qwen3 Abliterated FP16 Collection (4B + 8B)

This repository is a unified model card for FP16 single-file safetensors conversions of Qwen3 abliterated variants. It is maintained as a combined release line for both 4B and 8B.

Model Scope

Variant	Upstream Base	Format	Status
4B	huihui-ai/Qwen3-4B-abliterated	Single .safetensors in FP16	Available
8B	huihui-ai/Qwen3-8B-abliterated	Single .safetensors in FP16	To be published in this repository

Repository Files

• qwen3_4b_abliterated_fp16_converted.safetensors (currently available)
• 8B FP16 single-file safetensors will be added to the same repository.

Conversion Workflow

4B conversion

1. Download sharded safetensors from the 4B source repository.
2. Merge shards into one tensor file.
3. Convert all weights to FP16.
4. Save as a single safetensors artifact.

8B conversion

1. Download sharded safetensors from huihui-ai/Qwen3-8B-abliterated.
2. Merge all shard files into one tensor file.
3. Convert all weights to FP16.
4. Publish the single-file FP16 artifact in this repository.

Usage (Transformers)

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model_id = "ussoewwin/qwen3_4b_abliterated_fp16"

model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.float16,
    device_map="auto",
    trust_remote_code=True,
)
tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)

messages = [{"role": "user", "content": "Hello, how are you?"}]
input_ids = tokenizer.apply_chat_template(
    messages,
    tokenize=True,
    add_generation_prompt=True,
    return_tensors="pt",
).to(model.device)

outputs = model.generate(input_ids, max_new_tokens=512)
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(response)

Important Usage Notes

• This model family has reduced safety filtering compared to standard aligned chat checkpoints.
• Validate generated text before production or public deployment.
• Use in compliance with local laws, platform policies, and your own risk controls.

License

This repository follows Apache 2.0. Please also follow the upstream licenses and terms for each source model.

Acknowledgments

• 4B source: huihui-ai/Qwen3-4B-abliterated
• 8B source: huihui-ai/Qwen3-8B-abliterated
• Original base family: Qwen/Qwen3

Citation

If you use these models, please cite the original Qwen3 work:

@misc{qwen3,
  title={Qwen3: A Large-Scale Multilingual Language Model},
  author={Qwen Team},
  year={2024},
  howpublished={\url{https://github.com/QwenLM/Qwen3}}
}