Yi Su
- [x] I have marked all applicable categories: + [ ] exception-raising fix + [ ] algorithm implementation fix + [ ] documentation modification + [x] new feature - [x]...
- [x] I have marked all applicable categories: + [ ] exception-raising bug + [ ] RL algorithm bug + [ ] documentation request (i.e. "X is missing from the...
- [x] I have marked all applicable categories: + [ ] exception-raising bug + [ ] RL algorithm bug + [ ] documentation request (i.e. "X is missing from the...
- [x] I have marked all applicable categories: + [ ] exception-raising bug + [ ] RL algorithm bug + [ ] documentation request (i.e. "X is missing from the...
**Describe the feature** I was able to run `python tests/test_BERTNLU-RuleDST-RulePolicy-TemplateNLG.py` and got the same results as in README. Thanks a lot! However, as soon as I tried to run the...
I wonder why there isn't any GRU-based model implemented here. Reading [your paper](http://lxie.nwpu-aslp.org/papers/2020ICASSP_HJY.pdf), especially Fig. 1-4, I have the impression that GRU-based model performed better than TCN-based counterpart. I understand...
- [x] I have marked all applicable categories: + [ ] exception-raising bug + [ ] RL algorithm bug + [ ] documentation request (i.e. "X is missing from the...
Hi guys, Thank you so much for this awesome project! I really enjoyed it. Sorry for bringing this up again (#98). I saw @yoshisuga had tried it and even contributed...
Notes: - I left `validate()` intact; could use `get_dataloader()` as well and get rid of `get_batch()`; - Dependency on `transformers` can be eliminated by copying just the `LengthGroupedSampler`.
Informed by [this post](https://discuss.huggingface.co/t/fine-tuning-pre-training-tips/15367), I implemented the `group_by_length` feature in finetuning scripts and found on single V100 GPU `finetune/lora.py` LLaMA-7B with the same hyper parameters: - Without `group_by_length`: 8 hours,...