TimeLens2-27B-SFT

TimeLens2-27B-SFT is a video temporal grounding model based on Qwen/Qwen3.8-27B. Given a video and a text query, it predicts the relevant time intervals in seconds. This is the SFT version.

Paper ยท Code & Usage ยท Collection

Benchmark Results

Benchmark mIoU (%)
Charades-STA 62.26
ActivityNet Captions 64.12
QVHighlights 73.94
VUE-TR 50.81
VUE-TR-v2 49.60
MomentSeeker 30.06
Ego4D-NLQ-v2 22.54
Average 50.48

The two VUE benchmarks use the available-video subsets. The average is the unweighted mean over the seven benchmarks.

For inference and evaluation, follow the official TimeLens2 repository and use ZhuYuhan/TimeLens2-27B-SFT as the model path.

Downloads last month
43
Safetensors
Model size
27B params
Tensor type
BF16
ยท
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for ZhuYuhan/TimeLens2-27B-SFT

Base model

Qwen/Qwen3.8-27B
Finetuned
(950)
this model

Paper for ZhuYuhan/TimeLens2-27B-SFT