bing

koalas is a great package. when I install the package, all requirements are as below: `pip install koalas==1.8.2 Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple Collecting koalas==1.8.2 Using cached https://pypi.tuna.tsinghua.edu.cn/packages/28/9a/d69cf12ea62116873b427e5843be8ae8431b18f2a0714d6f4eec3ee4cda6/koalas-1.8.2-py3-none-any.whl (390 kB) Requirement...

ResourceWarning: Enable tracemalloc to get the object allocation traceback

1

`./venv/lib/python3.7/site-packages/flashtext/keyword.py:148: ResourceWarning: unclosed current_dict = current_dict.setdefault(letter, {}) ResourceWarning: Enable tracemalloc to get the object allocation traceback` hi @vi3k6i5 , I'm not familiar with this. Could you please help me take...

### Reminder - [X] I have read the README and searched the existing issues. ### Reproduction ```python def _train_(rank, world_size, allocated_gpus, parameters, callbacks): """assume the task use gpu 6, 7...

For example, the sentence "Presidencia de la República - Mexico", the word "de" will be classified as "en", but if i change it to " de ", as add space...

### Reminder - [X] I have read the README and searched the existing issues. ### Reproduction https://github.com/hiyouga/LLaMA-Factory/blob/main/src/llmtuner/train/sft/workflow.py sft支持do_predict。看CustomSeq2SeqTrainer继承于Seq2SeqTrainer，Seq2SeqTrainer实现了predict方法。 但是 https://github.com/hiyouga/LLaMA-Factory/blob/main/src/llmtuner/train/orpo/workflow.py ，orpo还未支持do_predict。看CustomORPOTrainer继承于DPOTrainer，但是DPOTrainer并未实现predict方法。 我看了https://github.com/huggingface/trl/blob/main/trl/trainer/orpo_trainer.py#L800，也没有实现predict方法。 请问： 1，是因为本身DPOTrainer，orpo_trainer，理论上不适合实现predict方法吗？有什么原因，不实现predict方法吗？因为都是继承trainer，为何不实现呢？ 2，如果也可以实现predict方法，有什么参考吗？ 我也还在尝试，但发现有些难度。 3，另外我看DPOTrainer，orpo_trainer都有一个get_batch_samples，是需要使用这个方法吗？ 多谢~ ### Expected...

how to use chinese llama 7b?

2

hi @scofield7419 @waxnkw your repo helps me a lot. I plan to use chinese llama 7b, but still get some errors. do I need to change these scripts? lavis/models/blip2_models/blip2_llama.py lavis/models/blip2_models/modeling_llama.py...

for sklearn, if needing install it, the ImportError should write "pip install scikit-learn" instead of "pip install sklearn" now: File "/home/ubuntu/wubing/evaluate/src/evaluate/loading.py", line 265, in _download_additional_modules raise ImportError( ImportError: To be...