issues
search
facebookresearch
/
metaseq
Repo for external large-scale work
MIT License
6.45k
stars
723
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Questions Regarding OPT Model Output
#758
srhouyu
opened
3 weeks ago
0
QA about continue training on checkpoint
#757
robinzixuan
opened
3 weeks ago
0
Update README.md
#756
flavioferlin
opened
4 months ago
1
minor changes
#755
herrminchen
opened
4 months ago
1
How to load the checkpoints into a HF model?
#754
jmkuebler
opened
7 months ago
0
Fixes
#753
sriniiyer
closed
1 year ago
0
Broadcast json index file rather than loading on all ranks
#752
sriniiyer
opened
1 year ago
0
I change Num_head of OPT-1.3b,and it cause CUDA Error: IndexSelectLargeIndex,
#751
Gusicun
opened
1 year ago
0
enable text eval with non-cm3
#750
lilisierrayu
opened
1 year ago
0
allow hubutils to support generators argument
#749
ramakanth-pasunuru
closed
1 year ago
0
Weights/Code for CM3Leon
#748
tomvars
opened
1 year ago
2
adding weighted
#747
lilisierrayu
opened
1 year ago
1
Implement JSONLDataset level partitioning
#746
ArmenAg
closed
1 year ago
1
setup to pyproject
#745
zycalice
opened
1 year ago
0
Implement Per-Modality PPL Tracking
#744
ArmenAg
closed
1 year ago
0
Fix marmot format variations for cm3v2
#743
berniebear
closed
1 year ago
0
fix and add marmot support for cm3v2
#742
berniebear
closed
1 year ago
0
How can I pretrain an opt-model with the codes?
#741
Gusicun
opened
1 year ago
0
add no_break_image break mode: no partial image at the beginning of a…
#740
violet-zct
closed
1 year ago
3
[fix] Set upper limit for aim version
#739
alberttorosyan
opened
1 year ago
1
Process blocks when deploying OPT-1.3B with FasterTransformer
#738
zhixin612
closed
1 year ago
0
Access request for opt-175b
#737
AlexEzx
opened
1 year ago
1
feat: add soft distillation
#736
mattmazzola
opened
1 year ago
0
OPT在中文对话上表现如何呢?
#735
LiZhangMing
opened
1 year ago
0
feat: add unified reshard script
#734
mattmazzola
opened
1 year ago
0
feat: add inference and evaluation script with dataset transformations
#733
mattmazzola
opened
1 year ago
0
feat: add demonstration of Sphinx documentation system
#732
mattmazzola
opened
1 year ago
0
feat: add .devcontainer to standardize development environment setup
#731
mattmazzola
opened
1 year ago
0
Fix lint / tests
#730
suchenzang
opened
1 year ago
1
fix: ensure last checkpoint is always saved, refactor training stop conditions to be computed in single location
#729
mattmazzola
opened
1 year ago
0
fix: add support for wide characters when building index of dataset files
#728
mattmazzola
closed
1 year ago
2
Cm3 integration
#727
urielsinger
opened
1 year ago
2
Possible feature and bugfix contributions from Microsoft research team's fork of Metaseq
#726
mattmazzola
opened
1 year ago
4
train opt-125M from scratch
#725
emrecanacikgoz
opened
1 year ago
2
Enable post ckpt callback, support local symlink
#724
adampolyak
closed
1 year ago
0
Add support for top-k
#723
EricMichaelSmith
closed
1 year ago
0
Grammatical Error Correction (GEC) prompt for OPT-IML
#722
yulonglin
opened
1 year ago
0
upgrade flask
#721
zycalice
closed
1 year ago
0
load checkpoint failed when training with multi-nodes.
#720
GongZhengLi
closed
1 year ago
1
Add type
#719
zycalice
opened
1 year ago
0
Davides/add efficiency metrics
#718
davides
closed
1 year ago
0
Generation should stop after two new lines if that is the stop criteria
#717
Vidyaranya
opened
1 year ago
2
OPT and LLaMA
#716
Paolo07700
closed
1 year ago
1
Andy/drop mseq req from reshard
#715
andrewPoulton
closed
1 year ago
1
FSDP is incompatible with BF16
#714
weigao266
closed
1 year ago
4
Add type hints to all methods
#713
suchenzang
opened
1 year ago
0
Fix an issue with the 6.7B path
#712
tangbinh
closed
1 year ago
0
Confirm md5sums after running reshard_fsdp.py on OPT-175B #702
#711
ayeeyecorp
closed
1 year ago
3
Codeowners / README update
#710
suchenzang
closed
1 year ago
0
Ignore checkpoint_last restart with NFS
#709
suchenzang
closed
1 year ago
1
Next