| .. |
|
configs
|
2fa8e69b62
add new argument: tokenizer_name
|
1 gadu atpakaļ |
|
data
|
4913d3ad24
Add missing copyright header
|
1 gadu atpakaļ |
|
datasets
|
69db75d425
fix incorrect split of InstructionDataset
|
1 gadu atpakaļ |
|
inference
|
81984a9a44
Remove unnecessary spec format
|
1 gadu atpakaļ |
|
model_checkpointing
|
ce9501f22c
remove relative imports
|
1 gadu atpakaļ |
|
policies
|
ce9501f22c
remove relative imports
|
1 gadu atpakaļ |
|
tools
|
4913d3ad24
Add missing copyright header
|
1 gadu atpakaļ |
|
utils
|
6954b16b3b
only save training params on rank 0
|
1 gadu atpakaļ |
|
finetuning.py
|
f8183b96fe
use new tokenizer_name argument and resize embeddings if required
|
1 gadu atpakaļ |
Rahul A R
2fa8e69b62
add new argument: tokenizer_name