423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-12-15 19:30:36 +08:00
Code Issues Packages Projects Releases Wiki Activity

update get template

Browse Source 
Former-commit-id: 21ea0d0786f91c0bce79630963e66b815a6792a0
This commit is contained in:
hiyouga
2024-09-04 22:36:20 +08:00
parent 1a76d3e09c
commit 3756553c25
17 changed files with 57 additions and 56 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
src/llamafactory/train/pt/workflow.py
View File

@@ -20,7 +20,7 @@ from typing import TYPE_CHECKING, List, Optional
from transformers import DataCollatorForLanguageModeling
from ...data import get_dataset
from ...data import get_dataset, get_template_and_fix_tokenizer
from ...extras.ploting import plot_loss
from ...model import load_model, load_tokenizer
from ..trainer_utils import create_modelcard_and_push
@@ -42,7 +42,8 @@ def run_pt(
):
tokenizer_module = load_tokenizer(model_args)
tokenizer = tokenizer_module["tokenizer"]
dataset_module, _ = get_dataset(model_args, data_args, training_args, stage="pt", **tokenizer_module)
template = get_template_and_fix_tokenizer(tokenizer, data_args)
dataset_module = get_dataset(template, model_args, data_args, training_args, stage="pt", **tokenizer_module)
model = load_model(tokenizer, model_args, finetuning_args, training_args.do_train)
data_collator = DataCollatorForLanguageModeling(tokenizer=tokenizer, mlm=False)