Commits · master · general / dsml / emoUS-public

Browse files

Mar 08, 2023
- transformers>=4.20.1,<=4.26.1 · 617e0025
  zqwerty authored Mar 8, 2023
  
  617e0025
- transformers>=4.20.1,<=4.26.1 · b73fffe1
  zqwerty authored Mar 8, 2023
  
  b73fffe1
Feb 26, 2023
- update merge_predict_res.py of t5nlu, t5dst, t5nlg · ce1d389d
  zqwerty authored Feb 26, 2023
  
  ce1d389d
Feb 23, 2023
- update run_seq2seq.py for rename inferenced file and handle error of AutoTokenizer on UnifiedQA_v2 · 4530095d
  zqwerty authored Feb 23, 2023
  
  4530095d
Jan 11, 2023
- T5NLU,T5DST,T5NLG for CrossWOZ · 8417eb23
  zqwerty authored Jan 11, 2023
  
  8417eb23
Dec 27, 2022
- train bertnlu and t5nlu for all utterance, train t5nlg for system&all utterances · 1f40d3d7
  zqwerty authored Dec 27, 2022
  
  1f40d3d7
Dec 17, 2022
- fix spelling mistake · b4dcb01d
  mikolajkrzyminski authored Dec 17, 2022
  
  b4dcb01d
Dec 13, 2022
- update nlu interface: auto preprocess for jointbert; transform da output for t5nlu · 58f9a0d4
  zqwerty authored Dec 13, 2022
  
  58f9a0d4
Dec 05, 2022
- rm useless import · bfb44786
  zqwerty authored Dec 5, 2022
  
  bfb44786
- fix bug in t5dst, t5nlg, and scgpt interface · 17229c70
  zqwerty authored Dec 5, 2022
  
  17229c70
Dec 04, 2022
- update nlu model link · a959c56a
  zqwerty authored Dec 4, 2022
  
  a959c56a
Nov 29, 2022
- rm gpt base model · a92c3a00
  zqwerty authored Nov 29, 2022
  
  a92c3a00
- add T5 readme · 8013a6fa
  zqwerty authored Nov 29, 2022
  
  8013a6fa
Oct 19, 2022
- update create_data:DST only consider the first value in slots' value list · fc3c6349
  zqwerty authored Oct 19, 2022
  
  fc3c6349
Aug 02, 2022
- merge seen&unseen testset of wow · ae558fd5
  zqwerty authored Aug 2, 2022
  
  ae558fd5
- do not update ft data, but update evaluate & metric · 7a6522f0
  zqwerty authored Aug 2, 2022
  
  7a6522f0
- Revert "update ft data processing" · edbebd06
  zqwerty authored Aug 2, 2022
  edbebd06
Aug 01, 2022
- update ft data processing · df7e29a3
  zqwerty authored Aug 1, 2022
  
  df7e29a3
Jul 25, 2022
- write finetune result to csv · 700b6017
  zqwerty authored Jul 25, 2022
  
  700b6017
Jul 22, 2022
- fix bug in dataset_vanilla.py, add dataset_godel.py · b0d0851e
  zqwerty authored Jul 22, 2022
  
  b0d0851e
- update evaluate.py · e90344db
  zqwerty authored Jul 22, 2022
  
  e90344db
- change --adafactor to --optim adafactor · 3775f099
  zqwerty authored Jul 22, 2022
  
  3775f099
- add key2gen+noisy w/o rg script · 824068f8
  zqwerty authored Jul 22, 2022
  
  824068f8
Jul 21, 2022
- add finetune suite and update pre-training scripts · 5a1ed974
  zqwerty authored Jul 21, 2022
  
  5a1ed974
Jul 18, 2022
- extend dataset preprocessing time from 30 mins to 1 year · fb70aed4
  zqwerty authored Jul 18, 2022
  
  fb70aed4
- extend dataset preprocessing time from 30 mins to 1 year · f34083f9
  zqwerty authored Jul 18, 2022
  
  f34083f9
- multitask pre-training · ca6cf298
  zqwerty authored Jul 18, 2022
  
  ca6cf298
Jul 17, 2022
- pre-train t5 rg · e4872bae
  zqwerty authored Jul 17, 2022
  
  e4872bae
- store sen2gen trial · a9443416
  zqwerty authored Jul 17, 2022
  
  a9443416
- rm keywords_th_ratio, add keywords_loss_th args for keyword extraction · 1771c52b
  zqwerty authored Jul 17, 2022
  
  1771c52b
- rm get_word_loss.sh · 7570883d
  zqwerty authored Jul 17, 2022
  
  7570883d
- update lmloss2keywords to reduce memory cost · dafb50a2
  zqwerty authored Jul 17, 2022
  
  dafb50a2
Jul 15, 2022
- add __init__.py · 50e5f144
  zqwerty authored Jul 15, 2022
  
  50e5f144
- update get token loss method for multi-gpu inference · 76c1d32e
  zqwerty authored Jul 15, 2022
  
  76c1d32e
Jul 12, 2022
- replace HF seq2seq trainer to convlab seq2seq trainer to allow more config for... · 02bd56be
  zqwerty authored Jul 12, 2022
  02bd56be
- add retrieval augmented NLU · bd7a97b6
  zqwerty authored Jul 12, 2022
  
  bd7a97b6
Jul 08, 2022
- t5dst multitask · 61b6c665
  zqwerty authored Jul 8, 2022
  
  61b6c665
Jul 06, 2022
- nlu multitask exp · 06ed57f4
  zqwerty authored Jul 6, 2022
  
  06ed57f4
Jul 04, 2022
- nlg multitask exp · 2cb15bb2
  zqwerty authored Jul 4, 2022
  
  2cb15bb2
Jun 22, 2022
- update requirements.txt & setup.py, move tests/ to agent_examples/, move deploy to convlab/deploy · 3393a064
  zqwerty authored Jun 22, 2022
  
  3393a064