Usage
# pip install "transformers==4.51.3" "peft==0.20.0" 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-2fps-s0")
The evaluation and audit scripts of the code repository take --model <base> --adapter <this repository or folder>.
Training
GRPO with a temporal-IoU reward, at 2 fps on the same 2,400 training pairs as the placebo-controlled models.
Table with columns: Setting, Value| Setting | Value |
|---|
| KL coefficient | 0.04 |
| Samples per question | 8 |
| Questions per step | 4 |
| Warm-up steps | 20 |
| Training pairs | 2,400 from the Charades-STA training split |
| Optimizer steps | 600 |
| Learning rate | 1e-5 |
| Seed | 0 |
| LoRA | r=16, alpha=32, dropout 0.05, on q/k/v/o and gate/up/down projections |
| Video input | 2 fps, at most 768 frames |
| Prompt template | timer1_default |
| Trainer | code/29_grpo_native_iou.py |
Citation
Anonymous. Accuracy Is Not Evidence: Measuring and Training Evidence Dependence in Video Temporal Grounding. ICLR 2027 submission, under review.