Tristan Bandat

icon location Linz icon location Hobby programmer and AI student @ JKU.

Results 3 comments of Tristan Bandat

Significant Inference Time Discrepancy Between craft_mlt_25k.pth and CRAFT_clr_amp_29500.pth on Various CPUs

I have the same issue on an `Intel Core i7-10850H`. Own model is 10x slower. Quantification to `torch.qint8` (as in the original `get_detector()` function) does not help.

GPU Batch Inference Implementation for SAHI

> Also, am I correct to assume that for now only Ultralytics support is included? @vittorio-prodomo As far I'm concerned the `UltralyticsDetectionModel` class is also used for e.g. PyTorch models....

Freethreading status tracking issue

Is there already a release planed which includes this feature?