Shark-NLP DiffuSeq issues

Shark-NLP / DiffuSeq

[ICLR'23] DiffuSeq: Sequence to Sequence Text Generation with Diffusion Models

MIT License

707 stars 87 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Text simplification dataset

#85 MeshchaninovViacheslav opened 2 months ago
1
questions on source data

#84 louisefz opened 2 months ago
1
Understanding tT_loss

#83 orxh closed 3 months ago
2
'grad_norm' is NaN

#82 LikeStarting opened 4 months ago
3
Where is CommonsenseConversation/test.jsonl ? When I run train. sh and then run run_decode_solver. sh or run_decode. sh, I always can't find test.jsonl

#81 Humble2967738843 closed 4 months ago
2
train

#80 mattat0516 opened 4 months ago
0
DDPM

#79 liurob2000 opened 5 months ago
3
ddim sampling

#78 BIT-MJY closed 5 months ago
2
Implementation of using soft absorbing state in the forward process in training.

#77 siebeniris closed 5 months ago
1
A question about the loss in V2

#76 kingkingofall closed 6 months ago
0
Merging changes

#75 chiral-carbon closed 6 months ago
1
Machine Translation Task with DiffuSeq

#74 chiral-carbon opened 6 months ago
6
If there is any rule to modify the parameters

#73 zkzhou126 opened 7 months ago
1
i face some promble Dataset(2) in "text_datasets.py"

#72 decode12 opened 8 months ago
1
Could not find a version that satisfies the requirement torch==1.9.0+cu111

#71 phuocdn1115 opened 8 months ago
0
a question about --local_rank

#70 chenghl666 opened 8 months ago
0
Incorrect self-BLEU Computation

#69 zzbuzzard opened 8 months ago
0
Version of many packages

#68 chooonsik opened 9 months ago
0
a few questions about the 'MBR' decoding strategy.

#67 xiaotingxuan closed 8 months ago
2
Try to train the model with another dataset, but get so many [UNK] token.

#66 Neochris closed 9 months ago
1
BERT parameter

#65 X-fxx opened 9 months ago
0
Padding during training results in a "Killed"

#64 MKSharaf opened 9 months ago
0
Issues with decoding and evaluation

#63 chiral-carbon opened 10 months ago
2
DiffuSeq-v2 checkpoint release

#62 zzbuzzard opened 10 months ago
2
diffuseq-v2: TypeError: load_state_dict() takes 1 positional argument but 3 were given

#61 zzbuzzard opened 10 months ago
3
Thank you very much for coming out with the follow-up work V2, which brings a very important role to diffusion language modeling !I'd like to ask you some questions. Thank you.

#60 CCCCCCCABC closed 9 months ago
3
vocab dic invalid

#59 MirDoch closed 5 months ago
1
Sucessfully training log

#58 skepsun closed 10 months ago
1
NaN probabilities for step_sample

#57 xiaotingxuan closed 10 months ago
4
AttributeError: 'DataParallel' object has no attribute 'get_embeds'

#56 CCRss opened 11 months ago
0
about the data preprocess

#55 FayeXXX opened 1 year ago
0
Question for extending paraphrasing to other Dataset

#54 lizaitang opened 1 year ago
0
Why is trained embedding orthogonal?

#53 Peacer68 opened 1 year ago
1
Size of the hidden dimension

#52 mainpyp closed 1 year ago
2
RoFormer model instead of BERT?

#51 mainpyp closed 1 year ago
2
Taken <Pad> as a regular token could make model only learn the <Pad> information?

#50 ylgao1 opened 1 year ago
12
Gradient accumulation

#49 mainpyp closed 1 year ago
1
Decoding only [UNK] token ?

#48 cankobanz closed 1 year ago
1
Typo? train.py

#47 golankai closed 1 year ago
1
NLL for q0-q2 is 0 but for q3 is >2

#46 mainpyp opened 1 year ago
5
q<n> metrics

#45 mainpyp closed 1 year ago
1
Question About top-p sampling

#44 xiaotingxuan opened 1 year ago
1
Create LICENSE

#43 summmeer closed 1 year ago
0
License

#42 O-Schilter closed 1 year ago
1
NCCL error

#41 johnnytam100 opened 1 year ago
0
Nothing generated from decode

#40 johnnytam100 opened 1 year ago
2
Fix data sampling in different ranks

#39 kdha0727 closed 1 year ago
0
Where decode "src" and "trg" ?

#38 cankobanz closed 1 year ago
1
the Multi-GPU training acutally duplicates data in each GPU ?

#37 Dawn-LX closed 1 year ago
19
Dataset(2) in "text_datasets.py"

#36 minorfox closed 1 year ago
4