refactor(nn): improve module naming and registration#433
Merged
chenghuaWang merged 7 commits intoUbiquitousLearning:v2from Sep 8, 2025
Merged
refactor(nn): improve module naming and registration#433chenghuaWang merged 7 commits intoUbiquitousLearning:v2from
chenghuaWang merged 7 commits intoUbiquitousLearning:v2from
Conversation
Collaborator
chenghuaWang
commented
Sep 8, 2025
chenghuaWang
commented
- Update module registration to handle empty absolute name correctly
- Fix absolute name setting for layers
- Add FIXME comment for embedding output data type
- Make Context constructor private for singleton pattern
- Add MLLM_TYPE_MAPPING for tensor type conversion - Implement ParamsDict class for storing and manipulating tensors - Add functions for loading PyTorch and safetensors models - Implement tensor conversion methods in ParamsDict
…dels - Add support for loading and storing MLLM V1 and V2 model files in pure Python - Implement new functions in convertor module for model file operations - Update mllm_type_mapping.py to include more data type mappings - Create __init__.py in convertor module for unified interface
- Add mllm-quantizer tool for model quantization - Add mllm-convertor tool for model format conversion - Update setup and task files to include new tools - Refactor model file handling for better tool integration
…nagement - Add new executable targets for Qwen-2VL model - Implement KVCache reordering to improve memory efficiency - Update model configuration and tokenization for Qwen-2VL - Refactor code to support new model and optimization techniques
- Update module registration to handle empty absolute name correctly - Fix absolute name setting for layers - Add FIXME comment for embedding output data type - Make Context constructor private for singleton pattern
chenghuaWang
requested review from
liang1232018,
oreomaker,
xumengwei and
yirongjie
as code owners
- Add `start` parameter to `reorderKVCache` function for more efficient cache maintenance - Set log level to error in LazyVLMQwen2VL and LazyVLMQwen2_5VL to reduce output noise - Add `setLogLevel` function to mllm API for better control over logging
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.