issues
search
bigscience-workshop
/
t-zero
Reproduce results and replicate training fo T0 (Multitask Prompted Training Enables Zero-Shot Task Generalization)
Apache License 2.0
456
stars
53
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
fine tuning on other LM models
#43
ShiJiawenwen
opened
1 year ago
1
"Rank classification" in evaluation for multiple choices
#42
yuchenlin
opened
1 year ago
1
Update setup.py
#41
lintangsutawika
closed
1 year ago
1
docs: demo, experiments and live inference API on Tiyaro
#40
ijonglin
opened
1 year ago
0
Truncation method
#39
gportill
closed
2 years ago
4
Prediction for multi-token multiple choice?
#38
AADeLucia
closed
2 years ago
5
Evaluation on ANLI fails with `ValueError`
#37
fepegar
closed
2 years ago
3
Support Prefix & prompt upgrades
#36
Muennighoff
closed
2 years ago
7
T0 (p=1) replicability
#35
tuhinjubcse
opened
2 years ago
1
Added compatibility for XNLI
#34
lintangsutawika
closed
2 years ago
4
[WIP] Support new accelerate api
#33
thomasw21
opened
2 years ago
0
Fix padding side
#32
thomasw21
closed
2 years ago
1
Missing dataset config names in P3
#31
ari9dam
opened
2 years ago
2
big-bench evaluation
#30
thesofakillers
closed
2 years ago
1
run_eval changes the results depending on the batch size
#29
kkawamu1
closed
2 years ago
2
Cap the number of examples to lenth of dataset if it is smaller than 100
#28
kkawamu1
closed
2 years ago
1
class DecoderModel(ModelBase) using padding token for prediction
#27
richardbaihe
closed
2 years ago
1
Training duration
#26
kevinpl07
closed
2 years ago
1
mesh tensorflow checkpoints?
#25
hamishivi
closed
2 years ago
2
Fix promptsource dependency
#24
thomasw21
closed
2 years ago
1
how to reproduce the result in offline environment , I dowload the sentence.model and checkpoint but, the sentencepiece.model can not be recognised by the t5~!!
#21
flyingwaters
closed
2 years ago
1
how to use this t-zero to train my own dataset? did not find the api . if add the dataset to the cache_data_dir.I do not want reproduce. I want to use this tool train my own datasets
#20
flyingwaters
opened
2 years ago
4
how
#19
flyingwaters
closed
2 years ago
0
Question about the number of multi-task learning T0?
#18
StevenTang1998
closed
2 years ago
2
Fix bug for evaluation
#17
zheng-yanan
closed
2 years ago
1
Update citation and checkpoint descriptions
#16
awebson
closed
2 years ago
1
Fix package_data path for datasets.csv in setup.py
#15
craffel
closed
2 years ago
1
Fix typo in prompt name
#14
Guitaricet
closed
2 years ago
1
ETA on pytorch training code?
#13
NtaylorOX
opened
2 years ago
4
How much GPU Memory needed to finetune bigscience T0_3B Model?
#12
rajab-mondal07
closed
2 years ago
1
Remove requirement.txt
#11
thomasw21
closed
2 years ago
1
Add seqio tasks
#10
thomasw21
closed
2 years ago
0
Setup package
#9
thomasw21
closed
2 years ago
1
Instructions for fine-tuning a trained t-zero
#8
louis030195
closed
2 years ago
2
reproducibility versus replicability
#7
valdanchev
opened
2 years ago
1
Integrate `promptsource.seqio_tasks` into this repo
#6
VictorSanh
closed
2 years ago
2
reproducing/replicating training
#5
VictorSanh
closed
2 years ago
0
Factorize code to allow generic models + add decoder version
#4
thomasw21
closed
2 years ago
0
Fine-tuning/few-shot eval support
#3
awebson
closed
2 years ago
5
Inference
#2
VictorSanh
closed
2 years ago
0
Reproduce main evaluation
#1
VictorSanh
closed
2 years ago
0