upgradio 6.13.0

upgradio 6.12.0

feat(ui): allow selecting best frame directly from preview gallery

perf: cap shared inference to 16 frames for temporal consistency

fix: restore temporal sampling batch size expected by xportrait

fix(inference): match sample_log batch_size to nSample to avoid conditioning mismatch

fix: align run_inference signature with gradio app arguments

fix: add target_resolution passthrough to run_inference

fix: normalize image ranges safely in xportrait preprocessing

fix: update gradio video component usage for v6

perf: remove subprocess and optimize video pipeline for ZeroGPU

perf: optimize inference pipeline (cache model, avoid source re-encoding)

more specific gradio event

upgrade huggingface-hub 0.33.5

upgradio 6.9.0

upgradio 5.49.1

upgradio 5.45.0

upgradio 5.44.1 and disable embedding

upgradio 5.44.0

upgradio 5.43.1

Update app.py

Update app.py

Zero GPU readiness

Update core/test_xportrait.py

upgrade triton to 3.1.0

Update requirements.txt

upgrade pytorch dependencies (2.5.1)

upgrade compatible tokenizers

set up to date hf packages

typo

Update requirements.txt

Update requirements.txt

Update requirements.txt

upgradio to 5.34.1

add a time tracker

fixes 500 error for some users (#1)

Update README.md

add short description

Update app.py

Rename assets/inference_appearance_pose_local_mm_uc5_source_image_by_driving_video_mix4_converted.mp4 to assets/inference_result.mp4

Upload inference_appearance_pose_local_mm_uc5_source_image_by_driving_video_mix4_converted.mp4

Update app.py

Update app.py

Update requirements.txt

Update app.py

Update app.py

move examples

Update app.py

Update app.py

Update app.py