Skip to content

Remove tokenizer creation from sft example script #4197

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Oct 6, 2025
Merged

Remove tokenizer creation from sft example script #4197

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
78e3b7d
Remove tokenizer from sft example script
sergiopaniego Oct 2, 2025
67f335a
Merge branch 'main' into fix-sft
sergiopaniego Oct 6, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 2 additions & 8 deletions trl/scripts/sft.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -68,7 +68,7 @@

from accelerate import logging
from datasets import load_dataset
from transformers import AutoConfig, AutoModelForCausalLM, AutoTokenizer
from transformers import AutoConfig, AutoModelForCausalLM
from transformers.models.auto.modeling_auto import MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING_NAMES

from trl import (
Expand All @@ -93,7 +93,7 @@

def main(script_args, training_args, model_args, dataset_args):
################
# Model init kwargs & Tokenizer
# Model init kwargs
################
model_kwargs = dict(
revision=model_args.model_revision,
Expand All @@ -118,11 +118,6 @@ def main(script_args, training_args, model_args, dataset_args):
else:
model = AutoModelForCausalLM.from_pretrained(model_args.model_name_or_path, **model_kwargs)

# Create tokenizer
tokenizer = AutoTokenizer.from_pretrained(
model_args.model_name_or_path, trust_remote_code=model_args.trust_remote_code, use_fast=True
)

# Load the dataset
if dataset_args.datasets and script_args.dataset_name:
logger.warning(
Expand All @@ -145,7 +140,6 @@ def main(script_args, training_args, model_args, dataset_args):
args=training_args,
train_dataset=dataset[script_args.dataset_train_split],
eval_dataset=dataset[script_args.dataset_test_split] if training_args.eval_strategy != "no" else None,
processing_class=tokenizer,
peft_config=get_peft_config(model_args),
)

Expand Down
Loading