richardwth

icon company University of Melbourne icon location Melbourne icon location I am a researcher interested in transformers in NLP, generative models, implicit and physics informed NN, and reinforcement learning.

Results 11 comments of richardwth

How to preserve sections?

In V3.0.6, this can be solved by changing the default argument `mark_headers=False` to `mark_headers=True` at [extract.Extractor.clearn_text](https://github.com/attardi/wikiextractor/blob/f0ca16c3e92983b9094b6f32526992fc3a678f8f/wikiextractor/extract.py#L948). Then headings start with #, e.g. "## Section 1".

Option to keep == Section == syntax around titles

In V3.0.6, this can be solved by changing the default argument `mark_headers=False` to `mark_headers=True` at [extract.Extractor.clearn_text](https://github.com/attardi/wikiextractor/blob/f0ca16c3e92983b9094b6f32526992fc3a678f8f/wikiextractor/extract.py#L948). Then headings start with \#, e.g. "## Section 1".

how to sample images with a trained model

Sorry for the late reply. If you check line 499 of DeepLearning/my_sngan.py, you may notice the method `eval_sampling` randomly generates image samples and save the generated images to a sprite...

Wrong observation's start date with the MultifolderWithCache chronic class and the l2rpn_wcci_2022 env

I also noticed this long ago. My fix: ``` from grid2op.Chronics import GridStateFromFileWithForecasts class MyGridStateFromFileWithForecasts(GridStateFromFileWithForecasts): """ This class fixes a bug that the current chronix datatime is fixed to start...

Error occurs while training with rllib ray PPO

Same here. The error message is ```AttributeError: type object 'LightSimBackend_l2rpn_wcci_2022' has no attribute 'process_grid2op_shunt_data'``` Can be solved by changing `process_grid2op_shunt_data` to `process_shunt_data`.

Error occurs while training with rllib ray PPO

> can you train and load the agent with rllib ray correctly? No. I overwrite ray's `trainer.save_checkpoint` and `trainer.load_checkpoint` so that they are more user-friendly. Now they output straightforwardly model...

Is env.parameters.NB_TIMESTEP_COOLDOWN_SUB 0 or 3?

I further checked other parameters...and found four differences `make("l2rpn_wcci_2022").parameters.__dict__` gives ``` 'NB_TIMESTEP_OVERFLOW_ALLOWED': 2, 'NB_TIMESTEP_RECONNECTION': 10, 'NB_TIMESTEP_COOLDOWN_LINE': 0, 'NB_TIMESTEP_COOLDOWN_SUB': 0, ``` `make("../L2RPN_wcci2022_starting_kit/input_data_local").parameters.__dict__` gives ``` 'NB_TIMESTEP_OVERFLOW_ALLOWED': 3, 'NB_TIMESTEP_RECONNECTION': 12, 'NB_TIMESTEP_COOLDOWN_LINE': 3, 'NB_TIMESTEP_COOLDOWN_SUB':...

check cyclic error

I came across the same problem every time it needs to prove several points on a circle, i.e., `? cyclic a b c d`. The problem cannot be avoided by...

请问有计划推出中文的模型吗

不好意思，没有这方面的计划。我们提的方法好像也不能直接应用在中文上。

What pre-trained models do you use to retrieve negative embeddings?

Sorry for this late reply. We used a pretrained RoBERTa-large to calculate the embeddings of all samples, which were then used to calculate the similarities for negative sampling. This step...