issues
search
bigscience-workshop
/
multilingual-modeling
BLOOM+1: Adapting BLOOM model to support a new unseen language
https://arxiv.org/abs/2212.09535
Apache License 2.0
66
stars
14
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
torch: composable-sft requirement
#46
sbmaruf
opened
1 year ago
0
Turkish language training script
#45
sbmaruf
opened
1 year ago
0
Update README.md
#44
Muennighoff
closed
1 year ago
0
[WIP] Add ROOTS + Oscar mixed data
#43
haileyschoelkopf
opened
1 year ago
0
Calculate Trainable Parameters Correctly for `extend` vocab strategy.
#42
yongzx
opened
2 years ago
0
Merge JZ
#41
yongzx
closed
2 years ago
0
[Bug, needs reproduction] [low priority] Weird loss behavior with `pfeiffer+inv` config
#40
haileyschoelkopf
closed
2 years ago
1
Implement (IA)^3 in T-Few
#39
yongzx
closed
1 year ago
1
[Lower Priority] Implement "Fisher Induced Sparse uncHanging (FISH) Mask"
#38
haileyschoelkopf
opened
2 years ago
0
[Low Prio] Implement Diff-Pruning
#37
yongzx
opened
2 years ago
0
Trainable Parameters in madx_clm_run.py are incorrect for `extend` strategy.
#36
yongzx
opened
2 years ago
0
Implement Ladder Side-Tuning
#35
haileyschoelkopf
opened
2 years ago
1
Implement "Adaptable Adapters"
#34
haileyschoelkopf
opened
2 years ago
2
remove assert False
#33
yongzx
closed
2 years ago
0
add last-layer finetuning for tasks
#32
yongzx
closed
2 years ago
1
Implement LoRA.
#31
yongzx
closed
2 years ago
1
Implement Prefix-Tuning.
#30
yongzx
opened
2 years ago
0
Package the code up neatly to pass to JZ.
#29
yongzx
opened
2 years ago
1
Composable SFT
#28
haileyschoelkopf
opened
2 years ago
12
[WIP] Refactor madx_run_clm.py
#27
haileyschoelkopf
opened
2 years ago
7
Make generation task (XLSUM) work for decoder-only model and follow the original setting.
#26
yongzx
opened
2 years ago
0
Control Extra Params (use Adapter 16x reduction size as control)
#25
yongzx
opened
2 years ago
0
Add support for only the final task-specific layers in eval script
#24
yongzx
closed
2 years ago
0
Implement “Composable Sparse Fine-tuning for Cross-Lingual Transfer”
#23
haileyschoelkopf
opened
2 years ago
0
Bitfit
#22
lintangsutawika
closed
2 years ago
5
BitFit Finetuning.
#21
yongzx
closed
2 years ago
2
Check extend-vocab functionality; clean up extend-vocab model training
#20
haileyschoelkopf
closed
2 years ago
0
Support Embedding Strategy: Extend Vocab
#19
yongzx
closed
2 years ago
1
Unable to train Burmese tokenizer for training_size = 100,000
#18
yongzx
closed
2 years ago
5
Inconsistent Evaluation Results
#17
yongzx
opened
2 years ago
0
scripts for wikiann
#16
yongzx
closed
2 years ago
0
Eval wikiann
#15
yongzx
closed
2 years ago
0
XNLI evaluation (for baseline)
#14
yongzx
closed
2 years ago
1
updated madx_run_clm,py
#13
vnikouliNLE
closed
2 years ago
1
Add XLSum evaluation / unify eval script
#12
haileyschoelkopf
opened
2 years ago
6
Ext exp
#11
vnikouliNLE
closed
2 years ago
5
update xnli evaluation
#10
vnikouliNLE
closed
2 years ago
0
Adapter Training - Frozen transformer.wpe.weight?
#9
yongzx
closed
2 years ago
1
Sentence retrieval eval
#8
vnikouliNLE
closed
2 years ago
0
Incrementally adding new languages to multilingual checkpoint
#7
vnikouliNLE
closed
2 years ago
2
Adding Language Post-Training: MAD-X Adapters
#6
yongzx
closed
2 years ago
1
Adding Language Post-Training: Evaluation Suite
#5
yongzx
closed
2 years ago
1
data loader for multiatis
#4
sbmaruf
closed
2 years ago
8
Exp-001: Finetune gpt-2 model with new tokenizer on fr
#3
yongzx
closed
2 years ago
1
Incrementally adding new languages to pre-trained models
#2
hadyelsahar
closed
2 years ago
9
Adding Language specific validation sets to deepspeed
#1
hadyelsahar
opened
2 years ago
4