Add error handling to caption and focus generation functions in app_v4.py for improved robustness 843c09d ethix commited on Jun 12
Update demo launch method in app_v4.py to use queue for improved performance de8cc78 ethix commited on Jun 12
Add T5EncoderModel and FluxControlNetPipeline initialization in app_v4.py, set max memory usage for ZeroGPU a2ebda3 ethix commited on Jun 12
Remove duration parameter from GPU decorator in load_warm_models function in app_v4.py 10d29a5 ethix commited on Jun 12
Remove sequential CPU offload from model loading and simplify return value in safe_model_load function f725d52 ethix commited on Jun 12
Update generate_caption function to return a placeholder message when no control image is provided, and initialize space in app_v4.py b971b27 ethix commited on Jun 12
Update model initialization in app_v4.py to use captioner instead of model 9e2232a ethix commited on Jun 12
Add GPU initialization function in app_v4.py to manage model and pipeline setup cc343e2 ethix commited on Jun 12
