search

malllabiisc / EmbedKGQA

ACL 2020: Improving Multi-hop Question Answering over Knowledge Graphs using Knowledge Base Embeddings
Apache License 2.0
415 stars 95 forks source link

Creating model... Downloading: 0%| | 438k/501M [01:00<72:03:39, 1.93kB/s] #100

Closed cx51666666 closed 2 years ago

cx51666666 commented 2 years ago

Creating model... Downloading: 0%| | 438k/501M [01:00<72:03:39, 1.93kB/s]

What file do I need to download? Where else can I download it? Thank you！

apoorvumang commented 2 years ago

can you provide more details about what command you used?

cx51666666 commented 2 years ago

Thank you for your reply. The path is EmbedKGQA/KGQA/RoBERTa. Command: ~/project/EmbedKGQA/KGQA/RoBERTa$ python3 main.py --mode train --relation_dim 200 --do_batch_norm 1 --gpu 2 --freeze 1 --batch_size 16 --validate_every 10 --hops webqsp_half --lr 0.00002 --entdrop 0.0 --reldrop 0.0 --scoredrop 0.0 --decay 1.0 --model ComplEx --patience 20 --ls 0.05 --l3_reg 0.001 --nb_epochs 200 --outfile half_fbwq

apoorvumang commented 2 years ago

What's probably happening is huggingface (transformers library) is trying to download RoBERTa checkpoints + tokenizers. Does it not download for you?

cx51666666 commented 2 years ago

Due to network reasons, the download speed of the model is very slow, and there will be connection failure. Where can I download it manually? What path should I put it in after downloading? Thanks！

apoorvumang commented 2 years ago

I'm not sure, I think it would be better to refer to Huggingface's documentation for this. The path would be wherever huggingface's cache is (nothing EmbedKGQA specific).