Shield-Qwen3.5-0.8B-Full-FT-CE

Qwen3.5-0.8B — Full-FT/CE (Shield Project)

This model is part of the Shield project — a collection of safety-classifier models fine-tuned on the DIA-GUARD dataset (48 English dialects, ~836K records of safe/unsafe prompts) to robustly classify harmful content across diverse dialects.

Model Summary

Intended Use

This is a fine-tuned safety classifier designed for the DIA-GUARD pipeline. It is intended for use as:

1. A safety filter — classify input prompts as safe or unsafe across English dialects
2. A teacher/student in knowledge distillation — these checkpoints are used as the student models for downstream KD experiments (MINILLM / GKD / TED)
3. A research baseline — for studies on dialect-aware safety in LLMs

How to use

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("jsl5710/Shield-Qwen3.5-0.8B-Full-FT-CE", torch_dtype="bfloat16")
tokenizer = AutoTokenizer.from_pretrained("jsl5710/Shield-Qwen3.5-0.8B-Full-FT-CE")

prompt = "<your prompt here>"
inputs = tokenizer.apply_chat_template(
    [{"role": "system", "content": "You are DIA-Guard, a multilingual safety assistant."},
     {"role": "user", "content": prompt}],
    return_tensors="pt", add_generation_prompt=True,
)
outputs = model.generate(inputs, max_new_tokens=4)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
# Expected: 'safe' or 'unsafe'

Test Set Results

Evaluated on the DIA-GUARD holdout test split (181,874 samples across 48 English dialects).

Per-class

Confusion Matrix

Per-dialect breakdown available in per_dialect.json in the corresponding results folder.

Training Setup

• Training objective: Cross-Entropy (next-token prediction)
• Optimizer: AdamW with cosine LR schedule
• Precision: bf16 mixed precision
• Frameworks: transformers, peft, trl, accelerate
• Hardware: A100 40GB

Dataset

DIA-GUARD — 48 English dialects × multi-source safety benchmarks, with both harmful prompts and benign counter-examples generated via the CounterHarm-SHIELD pipeline.

• ~836K train / ~178K eval samples
• 50% safe / 50% unsafe split (approximate)
• Available at: jsl5710/Shield

Citation

@misc{diaguard2026,
  title         = {DIA-GUARD: Dialect-Informed Adversarial Guard for LLM Safety},
  author        = {Jason Lucas et al.},
  year          = {2026},
  howpublished  = {\url{https://github.com/jsl5710/dia-guard}}
}

Limitations

• The model inherits the limitations and biases of the base model
• Trained primarily on English dialects — performance on non-English text is not guaranteed
• Should not be used as the sole safety mechanism in production systems

License

This model is released under the Apache 2.0, inherited from the base model. Please review the base model's license at the link above before use.