| .. |
|
configs
|
fa0a389f74
add max_step feature for training and eval
|
hai 1 ano |
|
data
|
4913d3ad24
Add missing copyright header
|
hai 1 ano |
|
datasets
|
69db75d425
fix incorrect split of InstructionDataset
|
hai 1 ano |
|
inference
|
f63ba19827
Fixing tokenizer used for llama 3. Changing quantization configs on safety_utils.
|
hai 1 ano |
|
model_checkpointing
|
ce9501f22c
remove relative imports
|
%!s(int64=2) %!d(string=hai) anos |
|
policies
|
ce9501f22c
remove relative imports
|
%!s(int64=2) %!d(string=hai) anos |
|
tools
|
4913d3ad24
Add missing copyright header
|
hai 1 ano |
|
utils
|
e6f69f84ad
add max_steps_reached to reduce redundancy
|
hai 1 ano |
|
finetuning.py
|
11f51db28c
adding the kbit prep in the code
|
hai 1 ano |
Beto
f63ba19827
Fixing tokenizer used for llama 3. Changing quantization configs on safety_utils.