evidence-audit-timer1-7b-pc-a025-s0
LoRA adapter for Time-R1-7B, trained with the placebo-controlled objective at α = 0.25.
Released with the paper Accuracy Is Not Evidence: Measuring and Training Evidence Dependence in Video Temporal Grounding (ICLR 2027 submission, under review). Code, data identifiers and evaluation scripts are in iclropen/evidence-audit-code.
It appears in Tables 6, 7, 10, 11, 13, 14, 16, 19 and 20 and in Figures 1, 2, 4 and 5 of the paper.
Files
| Folder | Contents | transformers |
|---|---|---|
| root | final adapter | 4.51 (tested 4.51.3, peft 0.20) |
transformers-v5/ |
the same weights, with module names for transformers 5 | 5.x (tested 5.12.1, peft 0.19.1 or later) |
Adapters saved under transformers 4.51 and 5 use different module names. Loading one under the other version raises no error but leaves the LoRA weights unused, so use the version given here. tools/check_lora_keys.py in the code repository checks that an adapter actually loaded.
The transformers 4.51 runs used gradient checkpointing, under which the LoRA weights of the vision tower received no gradient. They are zero, so these adapters change the language model only.
Usage
# pip install "transformers==4.51.3" "peft==0.20.0" qwen-vl-utils
import torch
from transformers import Qwen2_5_VLForConditionalGeneration, AutoProcessor
from peft import PeftModel
base = "Boshenxx/Time-R1-7B"
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-timer1-7b-pc-a025-s0")
# with transformers 5: PeftModel.from_pretrained(model, "iclropen/evidence-audit-timer1-7b-pc-a025-s0", subfolder="transformers-v5")
The evaluation and audit scripts of the code repository take --model <base> --adapter <this repository or folder>.
Training
The placebo-controlled objective of Section 4 of the paper. It keeps the answer likelihood, weighted by α, and adds two margins for each edit operator: the annotated answer must be more likely on the clean video, and on a video with a matched placebo edit outside the event, than on the video with the event edited. A hinge keeps the drop under the placebo edit within a bound set by the reference model.
| Setting | Value |
|---|---|
| Operator weights | Group-DRO |
| α | 0.25 |
| β | 0.1 |
| Placebo bound | placebo drop ≤ 1.25 × reference drop + 0.10 |
| Training pairs | 2,400 from the Charades-STA training split |
| Optimizer steps | 300 |
| Gradient accumulation | 8 |
| Learning rate | 5e-6 |
| Seed | 0 |
| LoRA | r=16, alpha=32, dropout 0.05, on q/k/v/o and gate/up/down projections |
| Edit operators | swap, mask, blur, tshift, shuffle (two per example) |
| Video input | 2 fps, at most 768 frames |
| Prompt template | timer1_default |
| Trainer | code/27_papo_train_qcdro_v3seln_r5.py |
Citation
Anonymous. Accuracy Is Not Evidence: Measuring and Training Evidence Dependence in Video Temporal Grounding. ICLR 2027 submission, under review.