Detected kernel version 5.4.0, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher.
09/17/2024 17:17:16 - WARNING - __main__ - Process rank: 0, device: cuda:0, n_gpu: 1, distributed training: False, 16-bits training: False
09/17/2024 17:17:16 - INFO - __main__ - Training/evaluation parameters DistillationTrainingArguments(
_n_gpu=1,
accelerator_config={'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False},
adafactor=False,
adam_beta1=0.9,
adam_beta2=0.999,
adam_epsilon=1e-08,
auto_find_batch_size=False,
batch_eval_metrics=False,
bf16=False,
bf16_full_eval=False,
data_seed=None,
dataloader_drop_last=False,
dataloader_num_workers=8,
dataloader_persistent_workers=False,
dataloader_pin_memory=True,
dataloader_prefetch_factor=None,
ddp_backend=None,
ddp_broadcast_buffers=None,
ddp_bucket_cap_mb=None,
ddp_find_unused_parameters=None,
ddp_timeout=7200,
debug=[],
deepspeed=None,
disable_tqdm=False,
dispatch_batches=None,
do_eval=True,
do_predict=False,
do_train=True,
dtype=bfloat16,
eval_accumulation_steps=None,
eval_delay=0,
eval_do_concat_batches=True,
eval_on_start=False,
eval_steps=1000.0,
eval_strategy=no,
eval_use_gather_object=False,
evaluation_strategy=None,
fp16=False,
fp16_backend=auto,
fp16_full_eval=False,
fp16_opt_level=O1,
freeze_encoder=True,
fsdp=[],
fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False},
fsdp_min_num_params=0,
fsdp_transformer_layer_cls_to_wrap=None,
full_determinism=False,
generation_config=None,
generation_max_length=None,
generation_num_beams=None,
gradient_accumulation_steps=1,
gradient_checkpointing=True,
gradient_checkpointing_kwargs=None,
greater_is_better=None,
group_by_length=False,
half_precision_backend=auto,
hub_always_push=False,
hub_model_id=None,
hub_private_repo=False,
hub_strategy=every_save,
hub_token=<HUB_TOKEN>,
ignore_data_skip=False,
include_inputs_for_metrics=False,
include_num_input_tokens_seen=False,
include_tokens_per_second=False,
jit_mode_eval=False,
kl_weight=1.0,
label_names=None,
label_smoothing_factor=0.0,
learning_rate=0.0001,
length_column_name=length,
load_best_model_at_end=False,
local_rank=0,
log_level=passive,
log_level_replica=warning,
log_on_each_node=True,
logging_dir=./runs/Sep17_17-17-13_22c57e4734ce,
logging_first_step=False,
logging_nan_inf_filter=True,
logging_steps=25,
logging_strategy=steps,
lr_scheduler_kwargs={},
lr_scheduler_type=constant_with_warmup,
max_grad_norm=1.0,
max_steps=5000,
metric_for_best_model=None,
mp_parameters=,
neftune_noise_alpha=None,
no_cuda=False,
num_train_epochs=3.0,
optim=adamw_torch,
optim_args=None,
optim_target_modules=None,
output_dir=./,
overwrite_output_dir=True,
past_index=-1,
per_device_eval_batch_size=32,
per_device_train_batch_size=32,
predict_with_generate=True,
prediction_loss_only=False,
push_to_hub=True,
push_to_hub_model_id=None,
push_to_hub_organization=None,
push_to_hub_token=<PUSH_TO_HUB_TOKEN>,
ray_scope=last,
remove_unused_columns=True,
report_to=['tensorboard'],
restore_callback_states_from_checkpoint=False,
resume_from_checkpoint=None,
run_name=./,
save_on_each_node=False,
save_only_model=False,
save_safetensors=True,
save_steps=1000,
save_strategy=steps,
save_total_limit=1,
seed=42,
skip_memory_metrics=True,
sortish_sampler=False,
split_batches=None,
temperature=2.0,
tf32=None,
torch_compile=False,
torch_compile_backend=None,
torch_compile_mode=None,
torch_empty_cache_steps=None,
torchdynamo=None,
tpu_metrics_debug=False,
tpu_num_cores=None,
use_cpu=False,
use_ipex=False,
use_legacy_prediction_loop=False,
use_mps_device=False,
warmup_ratio=0.0,
warmup_steps=50,
weight_decay=0.0,
)
/root/anaconda/envs/distil-whisper/lib/python3.9/site-packages/huggingface_hub/utils/_deprecation.py:131: FutureWarning: 'Repository' (from 'huggingface_hub.repository') is deprecated and will be removed from version '1.0'. Please prefer the http-based alternatives instead. Given its large adoption in legacy code, the complete removal is only planned on next major release.
For more details, please read https://huggingface.co/docs/huggingface_hub/concepts/git_vs_http.
  warnings.warn(warning_message, FutureWarning)
/root/distil-whisper-large-v3-ptbr/./ is already a clone of https://huggingface.co/freds0/distil-whisper-large-v3-ptbr. Make sure you pull the latest changes with `repo.git_pull()`.
09/17/2024 17:17:17 - WARNING - huggingface_hub.repository - /root/distil-whisper-large-v3-ptbr/./ is already a clone of https://huggingface.co/freds0/distil-whisper-large-v3-ptbr. Make sure you pull the latest changes with `repo.git_pull()`.

Combining datasets...:   0%|          | 0/2 [00:00<?, ?it/s]
Combining datasets...:   0%|          | 0/2 [00:04<?, ?it/s]
Traceback (most recent call last):
  File "/root/distil-whisper-large-v3-ptbr/run_distillation.py", line 1632, in <module>
    main()
  File "/root/distil-whisper-large-v3-ptbr/run_distillation.py", line 799, in main
    raw_datasets["train"] = load_multiple_datasets(
  File "/root/distil-whisper-large-v3-ptbr/run_distillation.py", line 594, in load_multiple_datasets
    dataset = dataset.cast_column("audio", datasets.features.Audio(sampling_rate))
  File "/root/anaconda/envs/distil-whisper/lib/python3.9/site-packages/datasets/fingerprint.py", line 442, in wrapper
    out = func(dataset, *args, **kwargs)
  File "/root/anaconda/envs/distil-whisper/lib/python3.9/site-packages/datasets/arrow_dataset.py", line 2096, in cast_column
    dataset._data = dataset._data.cast(dataset.features.arrow_schema)
  File "/root/anaconda/envs/distil-whisper/lib/python3.9/site-packages/datasets/table.py", line 1577, in cast
    table = table_cast(self.table, target_schema, *args, **kwargs)
  File "/root/anaconda/envs/distil-whisper/lib/python3.9/site-packages/datasets/table.py", line 2283, in table_cast
    return cast_table_to_schema(table, schema)
  File "/root/anaconda/envs/distil-whisper/lib/python3.9/site-packages/datasets/table.py", line 2237, in cast_table_to_schema
    raise CastError(
datasets.table.CastError: Couldn't cast
filename: struct<bytes: binary, path: string>
  child 0, bytes: binary
  child 1, path: string
duration: double
transcript: string
transcript_mms: string
levenshtein: double
client_id: int64
filesize: int64
num_words: int64
-- schema metadata --
huggingface: '{"info": {"features": {"filename": {"sampling_rate": 24000,' + 390
to
{'filename': Audio(sampling_rate=24000, mono=True, decode=True, id=None), 'duration': Value(dtype='float64', id=None), 'transcript': Value(dtype='string', id=None), 'transcript_mms': Value(dtype='string', id=None), 'levenshtein': Value(dtype='float64', id=None), 'client_id': Value(dtype='int64', id=None), 'filesize': Value(dtype='int64', id=None), 'num_words': Value(dtype='int64', id=None), 'audio': Audio(sampling_rate=16000, mono=True, decode=True, id=None)}
because column names don't match
Traceback (most recent call last):
  File "/root/anaconda/envs/distil-whisper/bin/accelerate", line 8, in <module>
    sys.exit(main())
  File "/root/anaconda/envs/distil-whisper/lib/python3.9/site-packages/accelerate/commands/accelerate_cli.py", line 48, in main
    args.func(args)
  File "/root/anaconda/envs/distil-whisper/lib/python3.9/site-packages/accelerate/commands/launch.py", line 1174, in launch_command
    simple_launcher(args)
  File "/root/anaconda/envs/distil-whisper/lib/python3.9/site-packages/accelerate/commands/launch.py", line 769, in simple_launcher
    raise subprocess.CalledProcessError(returncode=process.returncode, cmd=cmd)
subprocess.CalledProcessError: Command '['/root/anaconda/envs/distil-whisper/bin/python', 'run_distillation.py', '--model_name_or_path', './distil-large-v3-init', '--teacher_model_name_or_path', 'openai/whisper-large-v3', '--train_dataset_name', 'freds0/cml_tts_dataset_portuguese+freds0/cml_tts_dataset_portuguese', '--train_split_name', 'train+test', '--train_dataset_config_name', 'default+default', '--text_column_name', 'transcript+transcript', '--eval_dataset_name', 'freds0/cml_tts_dataset_portuguese', '--eval_text_column_name', 'transcript', '--eval_steps', '1000', '--save_steps', '1000', '--warmup_steps', '50', '--learning_rate', '0.0001', '--lr_scheduler_type', 'constant_with_warmup', '--timestamp_probability', '0.2', '--condition_on_prev_probability', '0.2', '--language', 'pl', '--task', 'transcribe', '--logging_steps', '25', '--save_total_limit', '1', '--max_steps', '5000', '--wer_threshold', '20', '--per_device_train_batch_size', '32', '--per_device_eval_batch_size', '32', '--dataloader_num_workers', '8', '--preprocessing_num_workers', '8', '--ddp_timeout', '7200', '--dtype', 'bfloat16', '--output_dir', './', '--do_train', '--do_eval', '--gradient_checkpointing', '--overwrite_output_dir', '--predict_with_generate', '--freeze_encoder', '--streaming', 'False', '--push_to_hub']' returned non-zero exit status 1.