run_sft.py

# flake8: noqa
# Copyright 2023 The HuggingFace Inc. team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from trl.commands.cli_utils import SFTScriptArguments, TrlParser


from datasets import load_dataset

from transformers import AutoTokenizer
import models

from trl import (
    ModelConfig,
    SFTConfig,
    SFTTrainer,
    get_peft_config,
    get_quantization_config,
    get_kbit_device_map,
)


if __name__ == "__main__":
    parser = TrlParser((SFTScriptArguments, SFTConfig, ModelConfig))
    args, training_args, model_config = parser.parse_args_and_config()

    ################
    # Model init kwargs & Tokenizer
    ################
    quantization_config = get_quantization_config(model_config)
    model_kwargs = dict(
        revision=model_config.model_revision,
        trust_remote_code=model_config.trust_remote_code,
        attn_implementation=model_config.attn_implementation,
        torch_dtype=model_config.torch_dtype,
        use_cache=False,
        device_map=get_kbit_device_map() if quantization_config is not None else None,
        quantization_config=quantization_config,
    )
    training_args.model_init_kwargs = model_kwargs
    tokenizer = AutoTokenizer.from_pretrained(
        model_config.model_name_or_path, trust_remote_code=model_config.trust_remote_code, use_fast=True
    )
    tokenizer.pad_token = tokenizer.eos_token

    ################
    # Dataset
    ################
    dataset = load_dataset(args.dataset_name)

    ################
    # Training
    ################
    trainer = SFTTrainer(
        model=model_config.model_name_or_path,
        args=training_args,
        train_dataset=dataset[args.dataset_train_split],
        eval_dataset=dataset[args.dataset_test_split],
        tokenizer=tokenizer,
        peft_config=get_peft_config(model_config),
    )

    trainer.train()
    trainer.save_model(training_args.output_dir)