wxp
wxp
Any idea of training on GPU, not TPU? I am using Nvidia's V100, for the base model, I can only set the `train_batch_size=32`. BTW, what learning rate did you use...
> your representation is so confusing. train_batch_size=48 means batch size per gpu? or global batch size. if you use global batch size 48 your learning rate should be scaled (48/4096)^0.5...
> > your representation is so confusing. train_batch_size=48 means batch size per gpu? or global batch size. if you use global batch size 48 your learning rate should be scaled...
> see LARGE BATCH OPTIMIZATION FOR DEEP LEARNING: > TRAINING BERT IN 76 MINUTES(https://arxiv.org/abs/1904.00962) > also you'd better specify warm up step I did another round `ALBERT BASE model` training....
> It turned out that the file which the script tries to load as a vocabulary in fact is saved SentencePiece model. The change in the lines 159-161 did the...
The trained model is uncased, so the returned value of `do_lower_case` in `create_tokenizer_from_hub_module()` is `True` But in `class FullTokenizer`, when `spm_model_file` is not None, the current code ignore the the...
socket.io error when using custom react as frontend and chainlit as backend on cloud, local it works
Hello @suman12345678 How did you fix the above issue? I encountered the same issue for my chainlit ui which is deployed in GCP
hello @gilfernandes , the above version doesn't work for chainlit "0.7.2". The "made with chainlit" footer was still there after setting the custom_css in the config.toml file. Did I miss...