mukobi Unadaptable-Foundation-Models issues

mukobi / Unadaptable-Foundation-Models

MIT License

3 stars 0 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Benchmarking Online

#60 m-walters closed 4 months ago
0
Some configs stuff

#59 m-walters closed 4 months ago
0
Get things running

#58 m-walters closed 4 months ago
0
SPAR Report

#57 mukobi opened 4 months ago
0
SPAR Poster

#56 mukobi opened 4 months ago
0
Couple things

#55 m-walters closed 4 months ago
2
wandb.config object has no attribute 'finetune'

#54 owen-yeung closed 4 months ago
1
Pad token fix, FT no saving

#53 owen-yeung closed 4 months ago
0
Check if model tokenizer has PAD token, if not set EOS as PAD token

#52 owen-yeung opened 4 months ago
0
Rename 'outputs' and 'results' directories to something less confusing (or remove them)

#51 owen-yeung opened 4 months ago
0
Updated fine_tune configs

#50 owen-yeung closed 4 months ago
0
run_fine_tune Training needs pad token/no batching

#49 owen-yeung closed 4 months ago
1
Pythia testing

#48 owen-yeung closed 4 months ago
0
Make configs for fine tuning args (lr, epochs, etc) and pass into TrainingArguments in fine_tuning.run_fine_tune

#47 owen-yeung closed 4 months ago
2
fixed get_hf_data

#46 owen-yeung closed 4 months ago
0
minor fixes

#45 owen-yeung closed 4 months ago
0
PR for Shin branch

#44 m-walters closed 5 months ago
1
Sweeps both local and cluster

#43 m-walters closed 5 months ago
0
Tidy up imports

#42 m-walters closed 5 months ago
0
Create wandb team

#41 mukobi closed 5 months ago
0
Implement `get_unadapt_method`

#40 mukobi closed 4 months ago
1
utils.check_base_results_saved

#39 mukobi closed 4 months ago
2
Figure out how to store hyperparams in the repo along with base model results and write in readme

#38 mukobi opened 5 months ago
0
Implement metric.calculate_unadaptability_metrics

#37 ShinnosukeUesaka closed 5 months ago
1
Test pretrain_score once model.py is merged

#36 owen-yeung closed 5 months ago
1
implement wandb logging in run_fine_tune

#35 owen-yeung closed 5 months ago
0
implement run_fine_tune configs

#34 owen-yeung closed 5 months ago
0
run_fine_tune implemented

#33 owen-yeung closed 5 months ago
0
Fine_tune.run_fine_tune returns losses of training set rather than validation set

#32 ShinnosukeUesaka closed 5 months ago
2
Init config validation and Channels

#31 m-walters closed 5 months ago
0
Model perturbation unadapt method

#30 m-walters closed 5 months ago
1
Create batch script to run sweep agents in batch

#29 mukobi closed 5 months ago
0
Implement unadapt/model perturbations

#28 mukobi opened 5 months ago
0
Make config and test main.py for Pythia-14M

#27 mukobi closed 4 months ago
2
Implement zeroth order methods (MeZO) and test on tiny LLM

#26 mukobi opened 5 months ago
0
Write up entropy method in project overview

#25 mukobi opened 5 months ago
0
FIM Loss calculation bug

#24 m-walters closed 4 months ago
1
fix integrations

#23 ShinnosukeUesaka closed 5 months ago
4
Write readme

#22 owen-yeung closed 4 months ago
1
Impelement fine_tune.py

#21 owen-yeung closed 5 months ago
1
pretrain_score.run_benchmark

#20 owen-yeung closed 5 months ago
0
hf data loading implemented

#19 owen-yeung closed 5 months ago
0
Add model loading

#18 Rohan138 closed 5 months ago
2
main.py initial skeleton

#17 mukobi closed 5 months ago
0
Create cli tool, settings, and slurm utils

#16 m-walters closed 6 months ago
0
Readme contradiction

#15 m-walters closed 5 months ago
3
Write `utils.py`

#14 m-walters closed 5 months ago
1
Write `cli.py` util

#13 m-walters closed 5 months ago
0
Write `run_job.sh`

#12 m-walters closed 5 months ago
0
Write skeleton of `main.py`

#11 m-walters closed 5 months ago
0