issues
search
Modalities
/
modalities
A framework for training multimodal foundation models.
MIT License
39
stars
3
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Calculate total_steps for a scheduler dynamically.
#210
lllAlexanderlll
opened
38 minutes ago
0
feat: integrated pytorch swiglu and pytorch flash attention
#209
mali-git
closed
6 hours ago
0
Add special tokens and resize embedding layer
#208
lllAlexanderlll
opened
7 hours ago
0
Lora
#207
mrudat-iais
opened
8 hours ago
1
Fsdp 2.0 integration
#206
le1nux
opened
16 hours ago
0
Fix CPU Tests with Github Actions for torch>=2.4
#205
flxst
closed
1 day ago
0
Time logging checkpointing
#204
fromm-m
opened
3 days ago
0
Failing CPU Tests with Github Actions
#203
flxst
closed
1 day ago
0
feat: log model initialization & training start/end
#202
mali-git
closed
3 days ago
0
fix: improved readability by printing only on rank_0
#201
fromm-m
closed
3 days ago
0
Permanently fix pydantic warnings about model_ namespace
#200
flxst
opened
3 days ago
0
Fix pydantic warnings about model_ namespace
#199
flxst
closed
3 days ago
0
feat: moved print_rank_0 function towards a better fitting place
#198
fromm-m
closed
3 days ago
0
Feat/deferred init
#197
le1nux
closed
1 day ago
1
Instruction-tuning Support
#196
lllAlexanderlll
opened
1 week ago
2
Fix broken tests in main.
#195
lllAlexanderlll
closed
6 days ago
0
Wandb self contained offline logs
#194
le1nux
opened
1 week ago
0
SFT sample generator
#193
rrutmann
closed
1 week ago
0
SFT: Implement loss masking
#192
lllAlexanderlll
closed
1 week ago
0
Fix failing gpu tests
#191
le1nux
opened
2 weeks ago
0
Fix log only with global rank 0
#190
le1nux
closed
2 weeks ago
0
Refactor: Add Banner
#189
mali-git
closed
2 weeks ago
0
Refactor: Improve Readme
#188
mali-git
closed
2 weeks ago
0
Best-fit Packing
#187
lllAlexanderlll
opened
2 weeks ago
0
Fix: Computation of Total Number of Parameters
#186
mali-git
closed
3 weeks ago
0
Include mamba state space model dependency
#185
fromm-m
opened
3 weeks ago
0
Improve checkpoint conversion tests
#184
fromm-m
opened
3 weeks ago
0
Refactor the way we log the config and number of parameters
#183
le1nux
opened
3 weeks ago
0
bugfix: removed packaging from pyproject
#182
fromm-m
closed
3 weeks ago
0
feat: Unique Experiment ID
#181
mali-git
closed
3 weeks ago
0
Dataloader with fixed size
#180
le1nux
closed
2 weeks ago
0
Feat: Update Experiment ID
#179
mali-git
opened
3 weeks ago
0
Bring example configs up-to-date and remove legacy configs
#178
le1nux
opened
3 weeks ago
0
Packaging dependency not correctly installed during test pipeline execution
#177
le1nux
closed
3 weeks ago
0
Failing Test: No. Parameters Per Initialization Group
#176
mali-git
closed
2 days ago
1
Configurable Training Length
#175
flxst
opened
1 month ago
0
Support and test python versions 3.10 and 3.11
#174
flxst
closed
3 weeks ago
0
Supported Python Versions
#173
flxst
closed
3 weeks ago
0
Pull Request & Issue Templates
#172
flxst
closed
3 weeks ago
0
Fix getting started example
#171
flxst
closed
1 month ago
0
SwiGLU naming of projection matrices
#170
le1nux
closed
3 weeks ago
2
feat: log config
#169
mali-git
closed
3 weeks ago
0
Draft: Feat/initialization component
#168
le1nux
closed
1 month ago
0
Improve the dataset inheritance and class naming
#167
le1nux
opened
1 month ago
1
Fix getting started example
#166
flxst
closed
1 month ago
0
Limited and potentially incorrect weight initialization for CoCa model
#165
flxst
closed
1 month ago
1
Fix/dataset index: Index values were faulty when indexing the original samples instead of blocks.
#164
le1nux
closed
1 month ago
2
Bug: Dataset implementation does not
#163
le1nux
closed
3 weeks ago
0
Disable Flash Attention for inference
#162
rrutmann
opened
1 month ago
1
Feat: Various Configurable Initializations
#161
flxst
closed
3 weeks ago
0
Next