scripts/vlm/qwen25vl_finetune.py

# Copyright (c) 2025, NVIDIA CORPORATION.  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.

"""
# qwen2.5-3B with mock dataset
torchrun --nproc_per_node=2 qwen25vl_finetune.py --devices=2 --tp_size=1 --data_type=mock
# qwen2.5-7B with mock dataset
torchrun --nproc_per_node=8 qwen25vl_finetune.py --devices=8 --tp_size=4 --data_type=mock --model_size=7B
# qwen2.5-3B with with real dataset
torchrun --nproc_per_node=2 /opt/NeMo/scripts/vlm/qwen25vl_finetune.py \
    --data_type=qwen25vl --data_path=/datasets/sharegpt_test2.json \
    --gbs=2 --mbs=1 --devices=2 --tp=1 --max_sequence_length=8192
"""

import argparse

import torch
from lightning.pytorch.loggers import WandbLogger
from megatron.core.optimizer import OptimizerConfig
from transformers import Qwen2Tokenizer
from transformers.models.qwen2_vl.image_processing_qwen2_vl import Qwen2VLImageProcessor

from nemo import lightning as nl
from nemo.collections import llm, vlm
from nemo.collections.vlm import Qwen2VLDataConfig
from nemo.collections.vlm.qwen2vl.model import Qwen25VLVisionConfig
from nemo.lightning.pytorch.optim import CosineAnnealingScheduler
from nemo.lightning.pytorch.optim.megatron import MegatronOptimizerModule
from nemo.utils.exp_manager import TimingCallback


def main(args):
    # pylint: disable=C0115,C0116,C0301

    # Global and micro batch sizes
    gbs = args.gbs
    mbs = args.mbs
    max_steps = args.max_steps

    SIZE_INFO_MAP = {
        "3B": {"hf_model_name": "Qwen/Qwen2.5-VL-3B-Instruct", "llmconfig_class": llm.Qwen25Config3B},
        "7B": {"hf_model_name": "Qwen/Qwen2.5-VL-7B-Instruct", "llmconfig_class": llm.Qwen25Config7B},
        "32B": {"hf_model_name": "Qwen/Qwen2.5-VL-32B-Instruct", "llmconfig_class": llm.Qwen25Config32B},
        "72B": {"hf_model_name": "Qwen/Qwen2.5-VL-72B-Instruct", "llmconfig_class": llm.Qwen25Config72B},
    }

    model_size = args.model_size
    hf_model_name, llm_config_class = (
        SIZE_INFO_MAP[model_size]["hf_model_name"],
        SIZE_INFO_MAP[model_size]["llmconfig_class"],
    )

    max_sequence_length = args.max_sequence_length
    tokenizer = Qwen2Tokenizer.from_pretrained(hf_model_name)

    if args.data_type == "qwen25vl":
        # Data configuration
        data_config = Qwen2VLDataConfig(
            image_folder=args.image_folder,
            video_folder=args.video_folder,
            conv_template="qwen2vl",
            image_process_mode="square",
        )
        image_processor = Qwen2VLImageProcessor()
        # Data module setup
        data = vlm.Qwen2VLPreloadedDataModule(
            paths=args.data_path,
            model_version="qwen25-vl",
            data_config=data_config,
            seq_length=max_sequence_length,
            decoder_seq_length=None,
            global_batch_size=gbs,
            micro_batch_size=mbs,
            tokenizer=tokenizer,
            image_processor=image_processor,
            num_workers=1,
        )
    elif args.data_type == "mock":
        image_processor = Qwen2VLImageProcessor()
        data = vlm.Qwen2VLMockDataModule(
            seq_length=max_sequence_length,
            global_batch_size=gbs,
            micro_batch_size=mbs,
            tokenizer=tokenizer,  # for mock data, we generate random token directly, here tokenizer could be none
            image_processor=image_processor,
            num_workers=1,
        )
    else:
        raise ValueError(f"Data type {args.data_type} is not supported")

    # Submodules configurations
    language_transformer_config = llm_config_class(seq_length=max_sequence_length)
    vision_transformer_config = Qwen25VLVisionConfig()
    vision_projection_config = vlm.MultimodalProjectorConfig(
        projector_type=args.projector_type,
        input_size=vision_transformer_config.hidden_size * (vision_transformer_config.spatial_merge_size**2),
        hidden_size=language_transformer_config.hidden_size,
        ffn_hidden_size=vision_transformer_config.hidden_size * (vision_transformer_config.spatial_merge_size**2),
    )

    # Qwen25VL model configuration
    qwen25vl_config = vlm.Qwen2VLConfig(
        language_transformer_config=language_transformer_config,
        vision_transformer_config=vision_transformer_config,
        vision_projection_config=vision_projection_config,
        language_model_from_pretrained=args.language_model_path,
        freeze_language_model=False,
        freeze_vision_model=True,
    )

    model = vlm.Qwen2VLModel(qwen25vl_config, model_version="qwen25-vl", tokenizer=data.tokenizer)

    from megatron.core.distributed import DistributedDataParallelConfig

    # Training strategy setup
    strategy = nl.MegatronStrategy(
        tensor_model_parallel_size=args.tp_size,
        pipeline_model_parallel_size=args.pp_size,
        encoder_pipeline_model_parallel_size=args.encoder_pp_size,
        pipeline_dtype=torch.bfloat16,
        sequence_parallel=args.enable_sp,
        ddp=DistributedDataParallelConfig(
            check_for_nan_in_grad=True,
            grad_reduce_in_fp32=True,
            overlap_grad_reduce=True,
            overlap_param_gather=True,
            average_in_collective=True,
        ),
    )

    # Checkpoint callback setup
    checkpoint_callback = nl.ModelCheckpoint(
        save_last=True,
        monitor="reduced_train_loss",
        save_top_k=2,
        every_n_train_steps=1000,
        dirpath=args.log_dir,
    )

    # Trainer setup
    trainer = nl.Trainer(
        num_nodes=args.num_nodes,
        devices=args.devices,
        max_steps=max_steps,
        accelerator="gpu",
        strategy=strategy,
        plugins=nl.MegatronMixedPrecision(precision="bf16-mixed"),
        callbacks=[checkpoint_callback, TimingCallback()],
        val_check_interval=gbs,
        limit_val_batches=0.0,
        log_every_n_steps=1,
        num_sanity_val_steps=0,
    )

    # Logger setup
    nemo_logger = nl.NeMoLogger(
        log_dir=args.log_dir,
        name=args.name,
        wandb=WandbLogger(project=args.wandb_project, name=args.name) if args.wandb_project is not None else None,
    )

    # Auto resume setup
    resume = nl.AutoResume(
        resume_if_exists=True,
        resume_ignore_no_checkpoint=True,
        resume_from_directory=args.log_dir,
        restore_config=nl.RestoreConfig(path=args.restore_path) if args.restore_path is not None else None,
    )

    # Optimizer and scheduler setup
    opt_config = OptimizerConfig(
        optimizer='adam',
        lr=args.lr,
        adam_beta1=0.9,
        adam_beta2=0.95,
        use_distributed_optimizer=True,
        bf16=True,
    )
    sched = CosineAnnealingScheduler(
        max_steps=trainer.max_steps,
        warmup_steps=0,
        constant_steps=1000,
        min_lr=1.0e-07,
    )
    opt = MegatronOptimizerModule(opt_config, sched)

    # PEFT setup
    if args.peft == 'lora':
        peft = vlm.peft.LoRA(
            target_modules=[
                "linear_qkv",
                "linear_proj",
                "linear_fc1",
                "linear_fc2",
            ]
        )
    else:
        peft = None

    llm.finetune(
        model=model,
        data=data,
        trainer=trainer,
        peft=peft,
        log=nemo_logger,
        optim=opt,
        resume=resume,
    )


if __name__ == "__main__":
    parser = argparse.ArgumentParser(description="QWEN2.5VL Model Training Script")

    # Argument parsing
    parser.add_argument("--data_type", type=str, required=False, default="mock", help="mock | qwen25vl")
    parser.add_argument("--data_path", type=str, required=False, default=None, help="Path to the dataset JSON file")
    parser.add_argument("--image_folder", type=str, required=False, default=None, help="Path to the image folder")
    parser.add_argument(
        "--video_folder",
        type=str,
        required=False,
        default=None,
        help="Path to the video folder, if not provided, use image_folder",
    )
    parser.add_argument(
        "--log_dir", type=str, required=False, default="/results", help="Directory for logging and checkpoints"
    )
    parser.add_argument(
        "--language_model_path", type=str, required=False, default=None, help="Path to the pretrained language model"
    )
    parser.add_argument(
        "--restore_path", type=str, required=False, default=None, help="Path to restore model from checkpoint"
    )
    parser.add_argument("--devices", type=int, required=False, default=1)
    parser.add_argument("--num_nodes", type=int, required=False, default=1)
    parser.add_argument("--max_steps", type=int, required=False, default=5190)
    parser.add_argument("--tp_size", type=int, required=False, default=1)
    parser.add_argument("--pp_size", type=int, required=False, default=1)
    parser.add_argument("--encoder_pp_size", type=int, required=False, default=0)
    parser.add_argument("--projector_type", type=str, required=False, default="mcore_mlp")
    parser.add_argument("--name", type=str, required=False, default="qwen25vl_pretrain")
    parser.add_argument("--peft", type=str, default='none', help="none | lora")
    parser.add_argument("--wandb_project", type=str, required=False, default=None)
    parser.add_argument("--gbs", type=int, required=False, default=64, help="Global batch size")
    parser.add_argument("--mbs", type=int, required=False, default=2, help="Micro batch size")
    parser.add_argument("--lr", type=float, required=False, default=2.0e-06, help="Learning rate")
    parser.add_argument('--enable_sp', action='store_true', help="enable sequence parallel")
    parser.add_argument(
        "--max_sequence_length", type=int, required=False, default=4096, help="Maximum sequence length"
    )
    parser.add_argument("--model_size", type=str, required=False, default="3B")

    args = parser.parse_args()
    main(args)