funasr2

游雁

2023-12-19 0e622e694e6cb4459955f1e5942a7c53349ce640

 funasr/bin/train.py

@@ -8,34 +8,30 @@
import hydra
from omegaconf import DictConfig, OmegaConf
from funasr.train_utils.set_all_random_seed import set_all_random_seed
# from funasr.model_class_factory1 import model_choices
from funasr.models.lora.utils import mark_only_lora_as_trainable
from funasr.optimizers import optim_choices
from funasr.schedulers import scheduler_choices
from funasr.optimizers import optim_classes
from funasr.schedulers import scheduler_classes
from funasr.train_utils.load_pretrained_model import load_pretrained_model
from funasr.train_utils.initialize import initialize
from funasr.datasets.fun_datasets.data_sampler import BatchSampler
# from funasr.tokenizer.build_tokenizer import build_tokenizer
# from funasr.tokenizer.token_id_converter import TokenIDConverter
from funasr.tokenizer.funtoken import build_tokenizer
from funasr.datasets.fun_datasets.dataset_jsonl import AudioDataset
# from funasr.tokenizer.funtoken import build_tokenizer
from funasr.train_utils.trainer import Trainer
# from funasr.utils.load_fr_py import load_class_from_path
from funasr.utils.dynamic_import import dynamic_import
import torch.distributed as dist
from torch.nn.parallel import DistributedDataParallel as DDP
from torch.distributed.fsdp import FullyShardedDataParallel as FSDP
from funasr.download.download_from_hub import download_model
from funasr.utils.register import registry_tables

@hydra.main(config_name=None, version_base=None)
def main_hydra(kwargs: DictConfig):
   import pdb; pdb.set_trace()
   if ":" in kwargs["model"]:
   assert "model" in kwargs
   if "model_conf" not in kwargs:
      logging.info("download models from model hub: {}".format(kwargs.get("model_hub", "ms")))
      kwargs = download_model(is_training=kwargs.get("is_training", True), **kwargs)
   
   import pdb;
   pdb.set_trace()

   main(**kwargs)


@@ -43,6 +39,7 @@
   # preprocess_config(kwargs)
   # import pdb; pdb.set_trace()
   # set random seed
   registry_tables.print_register_tables()
   set_all_random_seed(kwargs.get("seed", 0))
   torch.backends.cudnn.enabled = kwargs.get("cudnn_enabled", torch.backends.cudnn.enabled)
   torch.backends.cudnn.benchmark = kwargs.get("cudnn_benchmark", torch.backends.cudnn.benchmark)
@@ -56,31 +53,38 @@
      dist.init_process_group(backend=kwargs.get("backend", "nccl"), init_method='env://')
      torch.cuda.set_device(local_rank)
   
	
   # build_tokenizer
   tokenizer = build_tokenizer(
      token_type=kwargs.get("token_type", "char"),
      bpemodel=kwargs.get("bpemodel", None),
      delimiter=kwargs.get("delimiter", None),
      space_symbol=kwargs.get("space_symbol", "<space>"),
      non_linguistic_symbols=kwargs.get("non_linguistic_symbols", None),
      g2p_type=kwargs.get("g2p_type", None),
      token_list=kwargs.get("token_list", None),
      unk_symbol=kwargs.get("unk_symbol", "<unk>"),
   )
   # save config.yaml
   if (use_ddp or use_fsdp) and dist.get_rank() == 0 or not (use_ddp or use_fsdp) and local_rank == 0:
      os.makedirs(kwargs.get("output_dir", "./"), exist_ok=True)
      yaml_file = os.path.join(kwargs.get("output_dir", "./"), "config.yaml")
      OmegaConf.save(config=kwargs, f=yaml_file)
      logging.info("config.yaml is saved to: %s", yaml_file)

   tokenizer = kwargs.get("tokenizer", None)
   if tokenizer is not None:
      tokenizer_class = registry_tables.tokenizer_classes.get(tokenizer.lower())
      tokenizer = tokenizer_class(**kwargs["tokenizer_conf"])
      kwargs["tokenizer"] = tokenizer
	
   # build frontend if frontend is none None
   frontend = kwargs.get("frontend", None)
   if frontend is not None:
      frontend_class = registry_tables.frontend_classes.get(frontend.lower())
      frontend = frontend_class(**kwargs["frontend_conf"])
      kwargs["frontend"] = frontend
	
   # import pdb;
   # pdb.set_trace()
   # build model
   # model_class = model_choices.get_class(kwargs.get("model", "asr"))
   # model_class = load_class_from_path(kwargs.get("model").split(":"))
   model_class = dynamic_import(kwargs.get("model"))
   model_class = registry_tables.model_classes.get(kwargs["model"].lower())
   model = model_class(**kwargs, **kwargs["model_conf"], vocab_size=len(tokenizer.token_list))
   frontend = model.frontend



   # init_param
   init_param = kwargs.get("init_param", None)
   if init_param is not None:
      if not isinstance(init_param, Sequence):
      if not isinstance(init_param, (list, tuple)):
         init_param = (init_param,)
      logging.info("init_param is not None: %s", init_param)
      for p in init_param:
@@ -93,9 +97,8 @@
         )
   else:
      initialize(model, kwargs.get("init", "kaiming_normal"))
	
   # import pdb;
   # pdb.set_trace()


   # freeze_param
   freeze_param = kwargs.get("freeze_param", None)
   if freeze_param is not None:
@@ -122,33 +125,33 @@
      
   # optim
   optim = kwargs.get("optim", "adam")
   assert optim in optim_choices
   optim_class = optim_choices.get(optim)
   assert optim in optim_classes
   optim_class = optim_classes.get(optim)
   optim = optim_class(model.parameters(), **kwargs.get("optim_conf"))
   
   # scheduler
   scheduler = kwargs.get("scheduler", "warmuplr")
   assert scheduler in scheduler_choices
   scheduler_class = scheduler_choices.get(scheduler)
   assert scheduler in scheduler_classes
   scheduler_class = scheduler_classes.get(scheduler)
   scheduler = scheduler_class(optim, **kwargs.get("scheduler_conf"))


   # import pdb;
   # pdb.set_trace()
   # dataset
   dataset_tr = AudioDataset(kwargs.get("train_data_set_list"), frontend=frontend, tokenizer=tokenizer, **kwargs.get("dataset_conf"))
   dataset_class = registry_tables.dataset_classes.get(kwargs.get("dataset", "AudioDataset").lower())
   dataset_tr = dataset_class(kwargs.get("train_data_set_list"), frontend=frontend, tokenizer=tokenizer, **kwargs.get("dataset_conf"))

   # dataloader
   batch_sampler = BatchSampler(dataset_tr, **kwargs.get("dataset_conf"), **kwargs.get("dataset_conf").get("batch_conf"))
   batch_sampler = kwargs["dataset_conf"].get("batch_sampler", "DynamicBatchLocalShuffleSampler")
   batch_sampler_class = registry_tables.batch_sampler_classes.get(batch_sampler.lower())
   batch_sampler = batch_sampler_class(dataset_tr, **kwargs.get("dataset_conf"))
   dataloader_tr = torch.utils.data.DataLoader(dataset_tr,
                                               collate_fn=dataset_tr.collator,
                                               batch_sampler=batch_sampler,
                                               num_workers=kwargs.get("dataset_conf").get("num_workers", 4),
                                               pin_memory=True)
   
   if (use_ddp or use_fsdp) and dist.get_rank() == 0 or not (use_ddp or use_fsdp) and local_rank == 0:
      os.makedirs(kwargs.get("output_dir", "./"), exist_ok=True)
      yaml_file = os.path.join(kwargs.get("output_dir", "./"), "config.yaml")
      OmegaConf.save(config=kwargs, f=yaml_file)
      logging.info("config.yaml is saved to: %s", yaml_file)

   
   trainer = Trainer(
       model=model,