cocktailpeanut/LTX-Video-Playground

update

update

update

update

update

update

update

update

update

Update project page link in the UI for improved accessibility

better prompts

Reduce default number of inference steps from 40 to 30 for improved performance

Update project page badge color for improved visibility in the UI

Add project page link to the UI for improved accessibility

Add project page link and update default toggle values for prompt enhancement

Add title row for improved layout and organization in the UI

Add OpenAI integration and enhance prompt functionality for video generation

Update guidance scale range and add technical report links in the UI

Update default prompts for video generation to enhance user experience

Fix output index for preset change event handlers in video generation

Add debug print statements for video generation parameters

Change conditioning method to UNCONDITIONAL for video generation from text

Update preset options and slider default values for improved configuration

load unet in bfloat16

Add concurrency_id to video generation blocks for improved task management

Update iface.queue launch parameters to disable API exposure and hide API documentation for improved security

Enable queueing for txt2vid and img2vid presets in iface configuration for improved performance

Simplify iface.queue launch by removing unnecessary parameters for cleaner configuration

Remove max_thread_count from iface.launch configuration for streamlined setup

Update iface.queue configuration to include api_open, max_thread_count, and status_update_rate for enhanced control

Refactor app.py to launch iface.queue within a main guard for better script execution

Remove unnecessary pipeline.to("cpu") calls in video generation functions

Update app.py to reduce queue size to 64 and enable sharing in launch

Remove ssr_mode option from iface.launch for cleaner configuration

Add ssr_mode option to iface.launch for improved performance

Update app.py to enable queuing with a maximum size and hide API in launch

Enhance video generation error handling and memory management

force clean

remove intermediate tensors

Optimize video generation by adding torch.no_grad() context to reduce memory usage

Remove frame rate slider from advanced options in app.py

Add frame rate option to video generation functions and update UI labels

Update default resolution preset for improved performance

Rename video asset for consistency and clarity

Update example video assets to image formats for improved clarity

Add new example video assets and update existing prompts for clarity

Update example prompts and video assets for improved clarity and relevance

Refine negative prompt labels and update example video assets

Upload 6 files

Add example prompts and outputs for text-to-video and image-to-video generations