issues
search
nyu-mll
/
jiant-v1-legacy
The jiant toolkit for general-purpose text understanding models
MIT License
21
stars
9
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
[CLOSED] Fix target_train_data_fraction overriding pretrain_data_fraction
#1070
jeswan
closed
4 years ago
0
[CLOSED] Faster data fraction
#1069
jeswan
closed
4 years ago
6
zero-shot
#1068
jeswan
opened
4 years ago
2
[CLOSED] Backronym "jiant"?
#1067
jeswan
closed
4 years ago
2
[CLOSED] Results with pretrain_data_fraction args
#1066
jeswan
closed
4 years ago
2
[CLOSED] Google Cloud Platform Authenctication
#1065
jeswan
closed
4 years ago
2
[CLOSED] Increasing Length for each Example in MLM
#1064
jeswan
closed
4 years ago
2
[CLOSED] finetuning bert model
#1063
jeswan
closed
4 years ago
1
Transformer models do not use their corresponding pretrained poolers
#1062
jeswan
opened
4 years ago
1
[CLOSED] Adding Sentence Order Prediction
#1061
jeswan
closed
4 years ago
28
[DNM] XNLI and XLM-R
#1060
jeswan
opened
4 years ago
2
[CLOSED] Updating to Transformers v2.6.0
#1059
jeswan
closed
4 years ago
6
[CLOSED] Adding Sentence Order Prediction [WIP]
#1058
jeswan
closed
4 years ago
4
[CLOSED] Taskmaster scripts
#1057
jeswan
closed
4 years ago
3
Decouple evaluation tasks from target tasks
#1056
jeswan
opened
4 years ago
1
Speculative: Add support for automatic early stopping of poor runs
#1055
jeswan
opened
4 years ago
0
[CLOSED] analyzing the data after probing
#1054
jeswan
closed
4 years ago
3
special approval to use the Cove data (that's in the salesforce project)
#1053
jeswan
opened
4 years ago
8
Setup jiant for MultiRC
#1052
jeswan
opened
4 years ago
6
[CLOSED] problem with running the main.py file
#1051
jeswan
closed
4 years ago
3
[CLOSED] How to use the scripts in `/scripts`?
#1050
jeswan
closed
4 years ago
2
Unsupported tokenizer 'OpenAI.BPE'
#1049
jeswan
opened
4 years ago
7
Apex / FP16 support
#1048
jeswan
opened
4 years ago
3
[CLOSED] Fix FastText link in README
#1047
jeswan
closed
4 years ago
0
Set up T5
#1046
jeswan
opened
4 years ago
0
Word Embeddings for BLiMP testing
#1045
jeswan
opened
4 years ago
0
Apply Black formatting to files last modified before Black requirement
#1044
jeswan
opened
4 years ago
0
Revisit Integration of AllenNLP Models/Modules in jiant
#1043
jeswan
opened
4 years ago
0
[CLOSED] Does it possible to load and evaluate models using jiant other than the models implemented here?
#1042
jeswan
closed
4 years ago
4
Got RuntimeError: size mismatch when trying to train RoBERTa with SuperGLUE
#1041
jeswan
opened
4 years ago
11
Refactor load_tsv
#1040
jeswan
opened
4 years ago
0
[CLOSED] [WIP] Task load_data refactor
#1039
jeswan
closed
4 years ago
1
[CLOSED] Add test and documentation guidelines to CONTRIBUTING.md
#1038
jeswan
closed
4 years ago
0
[CLOSED] Perturbed
#1037
jeswan
closed
4 years ago
3
[CLOSED] Add CONTRIBUTING.md
#1036
jeswan
closed
4 years ago
0
[WIP] Trainer & __main__ interface refactor, some gpu related refactor
#1035
jeswan
opened
4 years ago
3
Fold RNNs, BoW into non-contextual word embeddings
#1034
jeswan
opened
4 years ago
0
Break up `models.py`
#1033
jeswan
opened
4 years ago
0
[CLOSED] Update refs in setup tutorial to v1.3.1
#1032
jeswan
closed
4 years ago
0
[CLOSED] fix mirrored pair input
#1031
jeswan
closed
4 years ago
0
[CLOSED] Adding Masked Language Modelling
#1030
jeswan
closed
4 years ago
14
[CLOSED] WIP: Extend edge probing fix to run in multi-GPU mode
#1029
jeswan
closed
4 years ago
0
[CLOSED] Add link to jiant 1.3.0 system overview paper in README
#1028
jeswan
closed
4 years ago
0
[CLOSED] Update author list
#1027
jeswan
closed
4 years ago
0
Phasing out ELMo support
#1026
jeswan
opened
4 years ago
1
[CLOSED] Fixing Edge-Probing after muti-GPU release
#1025
jeswan
closed
4 years ago
10
[CLOSED] Rohan
#1024
jeswan
closed
4 years ago
1
[CLOSED] Fixing cola inference script
#1023
jeswan
closed
4 years ago
0
[CLOSED] About nested configs
#1022
jeswan
closed
4 years ago
4
Performance issues on broadcoverage- and glue-diagnostic tasks
#1021
jeswan
opened
4 years ago
2
Previous
Next