Usage
# pip install "transformers>=5.12" "peft>=0.19.1" qwen-vl-utilsimport torchfrom transformers import Qwen2_5_VLForConditionalGeneration, AutoProcessorfrom peft import PeftModel base = "Qwen/Qwen2.5-VL-7B-Instruct"model = Qwen2_5_VLForConditionalGeneration.from_pretrained(base, torch_dtype=torch.bfloat16, device_map="auto")processor = AutoProcessor.from_pretrained(base)model = PeftModel.from_pretrained(model, "iclropen/evidence-audit-qwen7b-iougrpo-16frames-step700-s0")
The evaluation and audit scripts of the code repository take --model <base> --adapter <this repository or folder>.
Training
GRPO (TRL) with format and temporal-IoU rewards, on 16 frames per video. Its trainer is not part of the release.
Table with columns: Setting, Value| Setting | Value |
|---|
| Video input | 16 frames at 224 px |
| Optimizer steps | 700 (this checkpoint is the last step) |
| Questions per step / samples per question | 4 / 8 |
| Reward aggregation | sum, then normalise; no reward scaling |
| Gradient accumulation | 4 |
| Seed | 0 |
| LoRA | r=64, alpha=128, on q/k/v/o and gate/up/down projections |
| Training data | Charades-STA training split |
Citation
Anonymous. Accuracy Is Not Evidence: Measuring and Training Evidence Dependence in Video Temporal Grounding. ICLR 2027 submission, under review.