수고 많으십니다. 음성인식을 공부하고 있는 학생입니다. openspeech 사용 중 막힌 부분이 있어 질문드립니다.
hydra_lm_train.py 파일을 실행시키면 다음과 같은 오류가 발생했다고 합니다.
tokenizer가 없는 keyword라고 하는데 tokenizer를 지정 안 해주면 지정해주라고 하는데,,, 제 환경이 문제인 걸까요?
hydra_lm_train.py 역할이 무엇인가요? language model을 만드는 것인지 Acoustic model에 language model을 붙여서 새로운 모델을 만드는 것인지 궁금합니다. 이게 아니라면 acoustic model에 다른 language model을 붙이는 방법이 있을까요?
[2024-01-25 13:34:14,606][openspeech.utils][INFO] - Operating System : Windows 10
[2024-01-25 13:34:14,606][openspeech.utils][INFO] - Processor : Intel64 Family 6 Model 85 Stepping 4, GenuineIntel
[2024-01-25 13:34:14,607][openspeech.utils][INFO] - CUDA is available : False
[2024-01-25 13:34:14,607][openspeech.utils][INFO] - PyTorch version : 1.13.1+cpu
wandb: Currently logged in as: apg0001 (dguyanglab). Use wandb login --relogin to force relogin
wandb: wandb version 0.16.2 is available! To upgrade, please run:
wandb: $ pip install wandb --upgrade
wandb: Tracking run with wandb version 0.15.12
wandb: Run data is saved locally in C:\Users\lab1080\Desktop\openspeech\outputs\2024-01-25\13-34-14\wandb\run-20240125_133417-5gndlwut
wandb: Run wandb offline to turn off syncing.
wandb: Syncing run listen_attend_spell-ksponspeech
wandb: View project at https://wandb.ai/dguyanglab/listen_attend_spell-ksponspeech
wandb: View run at https://wandb.ai/dguyanglab/listen_attend_spell-ksponspeech/runs/5gndlwut
Traceback (most recent call last):
File "./openspeech_cli/hydra_lm_train.py", line 45, in hydra_main
data_module.setup(tokenizer=tokenizer)
TypeError: setup() got an unexpected keyword argument 'tokenizer'
❓ Questions & Help
수고 많으십니다. 음성인식을 공부하고 있는 학생입니다. openspeech 사용 중 막힌 부분이 있어 질문드립니다.
감사합니다.
Details
--입력 python ./openspeech_cli/hydra_lm_train.py dataset=ksponspeech dataset.dataset_path=C:\Users\lab1080\Desktop\openspeech\KsponSpeech dataset.test_dataset_path=C:\Users\lab1080\Desktop\openspeech\KsponSpeech_eval dataset.test_manifest_dir=C:\Users\lab1080\Desktop\openspeech\KsponSpeech_scripts dataset.manifest_file_path=C:\Users\lab1080\Desktop\openspeech\KSPONSPEECH_AUTO_MANIFEST model=listen_attend_spell lr_scheduler=warmup_reduce_lr_on_plateau trainer=gpu criterion=cross_entropy tokenizer=kspon_character
--출력 [2024-01-25 13:34:14,597][openspeech.utils][INFO] - dataset: dataset: ksponspeech dataset_path: C:\Users\lab1080\Desktop\openspeech\KsponSpeech test_dataset_path: C:\Users\lab1080\Desktop\openspeech\KsponSpeech_eval manifest_file_path: C:\Users\lab1080\Desktop\openspeech\KSPONSPEECH_AUTO_MANIFEST test_manifest_dir: C:\Users\lab1080\Desktop\openspeech\KsponSpeech_scripts preprocess_mode: phonetic criterion: criterion_name: cross_entropy reduction: mean lr_scheduler: lr: 0.0001 scheduler_name: warmup_reduce_lr_on_plateau lr_patience: 1 lr_factor: 0.3 peak_lr: 0.0001 init_lr: 1.0e-10 warmup_steps: 4000 model: model_name: listen_attend_spell num_encoder_layers: 3 num_decoder_layers: 2 hidden_state_dim: 512 encoder_dropout_p: 0.3 encoder_bidirectional: true rnn_type: lstm joint_ctc_attention: false max_length: 128 num_attention_heads: 1 decoder_dropout_p: 0.2 decoder_attn_mechanism: dot teacher_forcing_ratio: 1.0 optimizer: adam trainer: seed: 1 accelerator: dp accumulate_grad_batches: 1 num_workers: 4 batch_size: 32 check_val_every_n_epoch: 1 gradient_clip_val: 5.0 logger: wandb max_epochs: 20 save_checkpoint_n_steps: 10000 auto_scale_batch_size: binsearch sampler: else name: gpu device: gpu use_cuda: true auto_select_gpus: true tokenizer: sos_token:
eos_token:
pad_token:
blank_token:
encoding: utf-8
unit: kspon_character
vocab_path: ../../../aihub_labels.csv
[2024-01-25 13:34:14,606][openspeech.utils][INFO] - Operating System : Windows 10 [2024-01-25 13:34:14,606][openspeech.utils][INFO] - Processor : Intel64 Family 6 Model 85 Stepping 4, GenuineIntel [2024-01-25 13:34:14,607][openspeech.utils][INFO] - CUDA is available : False [2024-01-25 13:34:14,607][openspeech.utils][INFO] - PyTorch version : 1.13.1+cpu wandb: Currently logged in as: apg0001 (dguyanglab). Use wandb login --relogin to force relogin wandb: wandb version 0.16.2 is available! To upgrade, please run: wandb: $ pip install wandb --upgrade wandb: Tracking run with wandb version 0.15.12 wandb: Run data is saved locally in C:\Users\lab1080\Desktop\openspeech\outputs\2024-01-25\13-34-14\wandb\run-20240125_133417-5gndlwut wandb: Run wandb offline to turn off syncing. wandb: Syncing run listen_attend_spell-ksponspeech wandb: View project at https://wandb.ai/dguyanglab/listen_attend_spell-ksponspeech wandb: View run at https://wandb.ai/dguyanglab/listen_attend_spell-ksponspeech/runs/5gndlwut Traceback (most recent call last): File "./openspeech_cli/hydra_lm_train.py", line 45, in hydra_main data_module.setup(tokenizer=tokenizer) TypeError: setup() got an unexpected keyword argument 'tokenizer'