423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-10-15 16:18:10 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #434 from niuba/main

Browse Source 
add last_checkpoint support

Former-commit-id: b78d461f2826c194c332ead37825704c2cb8b910
This commit is contained in:
hoshi-hiyouga 2023-08-18 01:38:31 +08:00 committed by GitHub
commit f11c1ae562
1 changed files with 10 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
src/llmtuner/tuner/sft/workflow.py
View File

@ -1,4 +1,5 @@
# Inspired by: https://github.com/huggingface/transformers/blob/v4.29.2/examples/pytorch/summarization/run_summarization.py # Inspired by: https://github.com/huggingface/transformers/blob/v4.29.2/examples/pytorch/summarization/run_summarization.py
import os
from typing import TYPE_CHECKING, Optional, List from typing import TYPE_CHECKING, Optional, List
from transformers import DataCollatorForSeq2Seq from transformers import DataCollatorForSeq2Seq
@ -10,11 +11,14 @@ from llmtuner.extras.ploting import plot_loss
from llmtuner.tuner.core import load_model_and_tokenizer from llmtuner.tuner.core import load_model_and_tokenizer
from llmtuner.tuner.sft.metric import ComputeMetrics from llmtuner.tuner.sft.metric import ComputeMetrics
from llmtuner.tuner.sft.trainer import Seq2SeqPeftTrainer from llmtuner.tuner.sft.trainer import Seq2SeqPeftTrainer
from transformers.trainer_utils import get_last_checkpoint
from llmtuner.extras.logging import reset_logging, get_logger
if TYPE_CHECKING: if TYPE_CHECKING:
from transformers import Seq2SeqTrainingArguments, TrainerCallback from transformers import Seq2SeqTrainingArguments, TrainerCallback
from llmtuner.hparams import ModelArguments, DataArguments, FinetuningArguments, GeneratingArguments from llmtuner.hparams import ModelArguments, DataArguments, FinetuningArguments, GeneratingArguments
logger = get_logger(__name__)
def run_sft( def run_sft(
model_args: "ModelArguments", model_args: "ModelArguments",
@ -58,7 +62,12 @@ def run_sft(
# Training # Training
if training_args.do_train: if training_args.do_train:
train_result = trainer.train() checkpoint = None
if training_args.resume_from_checkpoint is not None:
checkpoint = training_args.resume_from_checkpoint
elif last_checkpoint is not None:
checkpoint = last_checkpoint
train_result = trainer.train(resume_from_checkpoint=checkpoint)
trainer.log_metrics("train", train_result.metrics) trainer.log_metrics("train", train_result.metrics)
trainer.save_metrics("train", train_result.metrics) trainer.save_metrics("train", train_result.metrics)
trainer.save_state() trainer.save_state()