from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "zhb12/Qwen3-0.6B-fulldata-v4-safety-epoch3"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    device_map="auto",
    dtype="auto",
)

messages = [{"role": "user", "content": "Hello!"}]
inputs = tokenizer.apply_chat_template(
    messages,
    add_generation_prompt=True,
    tokenize=True,
    return_dict=True,
    return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=128)
response = tokenizer.decode(
    outputs[0][inputs["input_ids"].shape[-1]:],
    skip_special_tokens=True,
)
print(response)

Use a Transformers version with Qwen3 support (transformers>=4.51.0).

Limitations

This checkpoint may produce inaccurate, unsafe, or biased content. Evaluate it for the intended domain before deployment. The training dataset and evaluation results have not been documented in this model card.

License

This derivative model is released under the Apache License 2.0, consistent with the base model.