issues about models
tanyue2019 opened this issue · comments
Hello, sorry to bother you! Is the generator model you used in "augment_otters_withpaths.py" the "checkpoints_6lendict_wcontains/model.ckpt" provided in the model folder of Google Drive? Why do I get the following error when I load this model?
Error(s) in loading state_dict for Generator:
Missing key(s) in state_dict: "gpt.h.0.attn.masked_bias", "gpt.h.1.attn.masked_bias", "gpt.h.2.attn.masked_bias", "gpt.h.3.attn.masked_bias", "gpt.h.4.attn.masked_bias", "gpt.h.5.attn.masked_bias", "gpt.h.6.attn.masked_bias", "gpt.h.7.attn.masked_bias", "gpt.h.8.attn.masked_bias", "gpt.h.9.attn.masked_bias", "gpt.h.10.attn.masked_bias", "gpt.h.11.attn.masked_bias"
For this script specifically, you need to use transformers Version: 2.8.0
There are multiple parts in this codebase and they are sometimes dependent on other existing repos. In those cases the scripts in this repo inherits the dependencies from those repos. Please feel free to create issues for any such dependency or version issue.
OK, thanks.！ I also wanted to try version 2.8.0, but it doesn't seem like it can be installed right now. Did you install it using pip?
You might need to downgrade your pytorch version
Ok I will try again. Thanks!
Hello , Can I ask what is the content of the folder "/alv2-tcfiltered0.8\u duppathremoved\u len12\u ppl1.3-vorgresp-checkpoint-500"
That is the model pretrained on data augmented using dailydialog. The folder can be found in the google drive.
Ok~Thanks!
Hi,Sorry to bother you again. Did you use different versions of transformers in different code files？