nlper

We have push the dataset to this respository, you can use `git clone https://github.com/scutcyr/CPED.git` to download the whole respository. The data can be found in `./data/CPED`. [https://github.com/scutcyr/CPED/tree/main/data/CPED](https://github.com/scutcyr/CPED/tree/main/data/CPED)

> Thanks! I also want to use the original/processed videos & audios, are they available now? Due to the copyright of the TV shows, we cannot publish the original audios...

By default, the speakers.txt generated by the `python generate_speakers4meld.py` has 30 speakers. Therefore, the `70+word2embed[speaker[j]]` would cause the `IndexError: column index (94) out of bounds`, because `70+30>90`! Thus, all the...

我们考虑在下一个版本的扁鹊模型同步开源数据集，当前仍在进一步优化迭代数据清洗。因为部分数据仍然可能涉及到隐私或者有害信息，我们尚不能公开当前版本的训练数据。

这个警告不影响推理和微调

我们后续将会提供基于FastAPI的后端部署代码。

Thanks for your question, the proactivehealthgpt_py38_.yml file for windows will be pushed soon!

> Thanks for your work !!! > > I used Win 10 platform and tried to build the conda env with your yml file mentioned in the README, but returns...

The following package is also need. ``` pip install streamlit_chat ```

微调的数据集v0.1已开源，详见：https://www.modelscope.cn/datasets/YIRONGCHEN/SoulChatCorpus