search

YuanGongND / ltu

Code, Dataset, and Pretrained Models for Audio and Speech Large Language Model "Listen, Think, and Understand".
367 stars 33 forks source link

vicuna_ltu model file missing #7

Open zengxijuan opened 9 months ago

zengxijuan commented 9 months ago

Hi, thank you for your wonderful work!

I've tried to run "finetune_toy.sh" following this:

prepare toy data and pretrained models

./prep_train.sh

run finetuning on the data

./finetune_toy.sh

But it occurs Error: "FileNotFoundError: [Errno 2] No such file or directory: '../../../pretrained_mdls/vicuna_ltu/pytorch_model-00001-of-00002.bin'", The directory downloaded is like

tree ../../../pretrained_mdls/vicuna_ltu/

../../../pretrained_mdls/vicuna_ltu/ |-- config.json |-- generation_config.json |-- pytorch_model-00003-of-00002.bin |-- pytorch_model.bin.index.json |-- pytorch_model_cavmae.bin.index.json |-- special_tokens_map.json |-- tokenizer.model `-- tokenizer_config.json

Where can I download this missing file?

YuanGongND commented 9 months ago

hi there,

thanks for your interest.

These files should be handled by ./prep_train.sh, you will need to run it before ./finetune_toy.sh.

If this does not solve the problem, it is possible that your region does not have dropbox service, in that case I highly recommend to use a VPN. To verify if it is a dropbox problem, can you try download the link?

https://github.com/YuanGongND/ltu/blob/4647269322473e8ce9fbb80a806dccf3643ab727/src/ltu/train_script/prep_train.sh#L22

We do provide a second option https://github.com/YuanGongND/ltu/tree/main#mirror-links, but that is a lot more complex (you need to download manually to correct path, and then build soft link etc)

-Yuan