Release 2.2.11: Add CogVideoX1.5-5B I2V Model Configuration #185
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add THUDM/CogVideoX1.5-5B-I2V model configuration with pipeline settings and enable I2V in random model selection.
benliang99
changed the title
…neration - Fix task assignment by using model group's task type instead of uninitialized model_name - Group models by task type to minimize prompt generator reloading - Add detailed logging for better monitoring - Remove duplicate random import
… handling - Refactor PromptGenerator to add load_vlm() and load_llm() for separate model loading - Use only load_llm() in DeepFloyd/IF NSFW retry loop to reduce memory usage - Add is_black_image utility to image_utils.py - Ensure LLM is loaded before prompt sanitization and clear GPU after - Minor code cleanup and comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Release 2.2.11 – I2V Expansion & Task-Specific Prompt Generation Fixes
Release Date: May 2, 2025
Version 2.2.11 introduces configuration support for the CogVideoX1.5-5B image-to-video model and fixing image and video prompt generation logic by correctly utilizing task types for annotation enhancement. It also prepares for a patch addressing legal NSFW filtering limitations in the DeepFloyd/IF model.
Updates
Model Changes
THUDM/CogVideoX1.5-5B-I2Vmodel with full pipeline configuration and support for I2V generationPrompt Generation Improvements
model_namerandommodule importKnown Issues and Upcoming Fix
Technical Details
Key changes include:
CogVideoX1.5-5B-I2Vinto theI2V_MODELSpipelineImpact
These changes strengthen the robustness of I2V generation by expanding model coverage and resolving incorrect task typing. Prompt generation is now more efficient and adaptable across task types.
Breaking Changes
model_name-based task detection