tackhwa
tackhwa
This is a Rabin-Karp Algorithm for Pattern Searching ### **Describe your change:** - [x] Add an algorithm? - [ ] Fix a bug or typo in an existing algorithm? -...
## Motivation Translate the documents of SlowFast & SlowOnly README document into Chinese ## Modification add `configs/detection/slowfast/README_zh-CN.md` add `configs/detection/slowonly/README_zh-CN.md` ## Checklist 1. Pre-commit or other linting tools should be used...
the configs file for mobilenetv3 is using "schedule_160k.py" , which is not fulfill training setting in [this link](https://github.com/tackhwa/mmsegmentation/blob/main/configs/mobilenet_v3/README.md). It should be 320k iterations instead of 160k. and the "Runner" is...
- Since the backbone in cofiguration file contain `_delete_=True`，it should be `model.merge()` instead of `model.update()`, or else the backbone will inherit other hyperparameter from `with read_base()` and cause error as...
- The Configs file had been tested, all the training can be run properly and the inference result is same as result in [this link](https://github.com/open-mmlab/mmsegmentation/tree/main/configs/maskformer) - Training for `maskformer_r50_d32_8xb2_160k_ade20k_512x512.py` ...
- The Configs file had been tested, all the training can be run properly and the inference result is same as result in [this link](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/mobilenet_v3/README.md) - Training for `mobilenet_v3_d8_s_lraspp_4xb4-320k_cityscapes_512x1024.py` ...
- new support feature is as title, google serper api had been refractor so that the response return by the function is having consistent formatting and will not throw an...
- tested on `MME` dataset  `observation`: It seems that Kosmos2 does not pre-train well with strong knowledge, resulting in poor performance on datasets that require extensive knowledge (testes with...