Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,18 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_381_4899e0da4615e883ad13.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_381_492597073098578f0605.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_381_67d9409a306e3614ec3f.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_381_7c47ba927ac118ffaacc.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_381_89ad32d31f3e70178cc1.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_381_cd644667186ae0518a3c.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_381_f7405ef7b645b3265477.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_762_09db3d4c95c46a3bae02.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_762_3349ee3cb71e3d789f00.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_762_63964662df5bc7731b7a.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_762_7ab4dededd2d3a052def.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_762_b8be5a72ea66619af007.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_762_d1c21fd700287f9f3109.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech[[:space:]]samples/eval_762_e8feadb7228a5b136ce8.wav filter=lfs diff=lfs merge=lfs -text
+wandb/run-20240513_205249-qaoje1x9/run-qaoje1x9.wandb filter=lfs diff=lfs merge=lfs -text

finetuning_concatenated_config.json

@@ -0,0 +1,54 @@
+{
+    "model_name_or_path": "parler-tts/parler_tts_mini_v0.1",
+    "feature_extractor_name": "parler-tts/dac_44khZ_8kbps",
+    "description_tokenizer_name": "parler-tts/parler_tts_mini_v0.1",
+    "prompt_tokenizer_name": "parler-tts/parler_tts_mini_v0.1",
+    "report_to": ["wandb"],
+    "overwrite_output_dir": true,
+    "train_dataset_name": "sanchit-gandhi/expresso-concatenated-half-normal+reach-vb/jenny_tts_dataset+sanchit-gandhi/libritts_r_test+sanchit-gandhi/libritts_r_test",
+    "train_metadata_dataset_name": "sanchit-gandhi/expresso-concatenated-half-normal-tags-mistral+ylacombe/jenny-tts-10k-tagged+parler-tts/libritts_r_tags_tagged_10k_generated+parler-tts/libritts_r_tags_tagged_10k_generated",
+    "train_dataset_config_name": "read+default+clean+other",
+    "train_split_name": "train+train[:20%]+test.clean+test.other",
+    "eval_dataset_name": "sanchit-gandhi/expresso-concatenated-half-normal+reach-vb/jenny_tts_dataset+sanchit-gandhi/libritts_r_test+sanchit-gandhi/libritts_r_test",
+    "eval_metadata_dataset_name": "sanchit-gandhi/expresso-concatenated-half-normal-tags-mistral+ylacombe/jenny-tts-10k-tagged+parler-tts/libritts_r_tags_tagged_10k_generated+parler-tts/libritts_r_tags_tagged_10k_generated",
+    "eval_dataset_config_name": "read+default+clean+other",
+    "eval_split_name": "train+train[:20%]+test.clean+test.other",
+    "max_eval_samples": 8,
+    "per_device_eval_batch_size": 16,
+    "target_audio_column_name": "audio",
+    "description_column_name": "text_description",
+    "prompt_column_name": "text",
+    "max_duration_in_seconds": 30.0,
+    "min_duration_in_seconds": 2.0,
+    "max_text_length": 400,
+    "preprocessing_num_workers": 2,
+    "do_train": true,
+    "num_train_epochs": 8,
+    "max_steps": -1,
+    "gradient_accumulation_steps": 8,
+    "gradient_checkpointing": true,
+    "per_device_train_batch_size": 16,
+    "learning_rate": 0.00008,
+    "adam_beta1": 0.9,
+    "adam_beta2": 0.99,
+    "weight_decay": 0.01,
+    "lr_scheduler_type": "cosine",
+    "warmup_steps": 250,
+    "logging_steps": 5,
+    "freeze_text_encoder": true,
+    "audio_encoder_per_device_batch_size": 4,
+    "dtype": "bfloat16",
+    "seed": 456,
+    "output_dir": "../output_dir_training_constant_concat/",
+    "temporary_save_to_disk": "../audio_code_tmp_constant_concat/",
+    "save_to_disk": "../tmp_dataset_audio_constant_concat/",
+    "dataloader_num_workers": 4,
+    "do_eval": true,
+    "predict_with_generate": true,
+    "include_inputs_for_metrics": true,
+    "save_strategy": "epoch",
+    "evaluation_strategy": "epoch",
+    "save_total_limit": 5,
+    "group_by_length": true
+}
+

run_parler_tts_training.py

@@ -0,0 +1,1763 @@
+#!/usr/bin/env python
+# coding=utf-8
+# Copyright 2024 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.
+
+""" Train Parler-TTS using 🤗 Accelerate"""
+
+import logging
+import os
+import re
+import shutil
+import sys
+import time
+from dataclasses import dataclass, field
+from datetime import timedelta
+from pathlib import Path
+from typing import Dict, List, Optional, Set, Union
+
+import datasets
+import evaluate
+import numpy as np
+import torch
+import transformers
+from accelerate import Accelerator
+from accelerate.utils import AutocastKwargs, InitProcessGroupKwargs, TorchDynamoPlugin, set_seed
+from accelerate.utils.memory import release_memory
+from datasets import Dataset, DatasetDict, IterableDataset, concatenate_datasets, interleave_datasets, load_dataset
+from huggingface_hub import Repository, create_repo
+from multiprocess import set_start_method
+from torch.utils.data import DataLoader
+from tqdm import tqdm
+from transformers import (
+    AutoFeatureExtractor,
+    AutoModel,
+    AutoProcessor,
+    AutoTokenizer,
+    HfArgumentParser,
+    Seq2SeqTrainingArguments,
+    pipeline,
+)
+from transformers.optimization import get_scheduler
+from transformers.trainer_pt_utils import LengthGroupedSampler
+from transformers.utils import send_example_telemetry
+from wandb import Audio
+
+from parler_tts import (
+    ParlerTTSConfig,
+    ParlerTTSForConditionalGeneration,
+    build_delay_pattern_mask,
+)
+
+
+logger = logging.getLogger(__name__)
+
+
+def list_field(default=None, metadata=None):
+    return field(default_factory=lambda: default, metadata=metadata)
+
+
+_RE_CHECKPOINT = re.compile(r"^checkpoint-(\d+)-epoch-(\d+)$")
+
+
+def get_last_checkpoint(folder):
+    content = os.listdir(folder)
+    checkpoints = [
+        path
+        for path in content
+        if _RE_CHECKPOINT.search(path) is not None and os.path.isdir(os.path.join(folder, path))
+    ]
+    if len(checkpoints) == 0:
+        return
+    return os.path.join(folder, max(checkpoints, key=lambda x: int(_RE_CHECKPOINT.search(x).groups()[0])))
+
+
+def sorted_checkpoints(output_dir=None, checkpoint_prefix="checkpoint") -> List[str]:
+    """Helper function to sort saved checkpoints from oldest to newest."""
+    ordering_and_checkpoint_path = []
+
+    glob_checkpoints = [str(x) for x in Path(output_dir).glob(f"{checkpoint_prefix}-*") if os.path.isdir(x)]
+
+    for path in glob_checkpoints:
+        regex_match = re.match(f".*{checkpoint_prefix}-([0-9]+)", path)
+        if regex_match is not None and regex_match.groups() is not None:
+            ordering_and_checkpoint_path.append((int(regex_match.groups()[0]), path))
+
+    checkpoints_sorted = sorted(ordering_and_checkpoint_path)
+    checkpoints_sorted = [checkpoint[1] for checkpoint in checkpoints_sorted]
+    return checkpoints_sorted
+
+
+def rotate_checkpoints(save_total_limit=None, output_dir=None, checkpoint_prefix="checkpoint") -> None:
+    """Helper function to delete old checkpoints."""
+    if save_total_limit is None or save_total_limit <= 0:
+        return
+    # Check if we should delete older checkpoint(s)
+    checkpoints_sorted = sorted_checkpoints(output_dir=output_dir, checkpoint_prefix=checkpoint_prefix)
+    if len(checkpoints_sorted) <= save_total_limit:
+        return
+
+    number_of_checkpoints_to_delete = max(0, len(checkpoints_sorted) - save_total_limit)
+    checkpoints_to_be_deleted = checkpoints_sorted[:number_of_checkpoints_to_delete]
+    for checkpoint in checkpoints_to_be_deleted:
+        logger.info(f"Deleting older checkpoint [{checkpoint}] due to args.save_total_limit")
+        shutil.rmtree(checkpoint, ignore_errors=True)
+
+
+def log_metric(
+    accelerator,
+    metrics: Dict,
+    train_time: float,
+    step: int,
+    epoch: int,
+    learning_rate: float = None,
+    prefix: str = "train",
+):
+    """Helper function to log all training/evaluation metrics with the correct prefixes and styling."""
+    log_metrics = {}
+    for k, v in metrics.items():
+        log_metrics[f"{prefix}/{k}"] = v
+    log_metrics[f"{prefix}/time"] = train_time
+    log_metrics[f"{prefix}/epoch"] = epoch
+    if learning_rate is not None:
+        log_metrics[f"{prefix}/learning_rate"] = learning_rate
+    accelerator.log(log_metrics, step=step)
+
+
+def log_pred(
+    accelerator,
+    pred_descriptions: List[str],
+    pred_prompts: List[str],
+    transcriptions: List[str],
+    audios: List[torch.Tensor],
+    sampling_rate: int,
+    step: int,
+    prefix: str = "eval",
+    num_lines: int = 200000,
+):
+    """Helper function to log target/predicted transcriptions to weights and biases (wandb)."""
+    if accelerator.is_main_process:
+        wandb_tracker = accelerator.get_tracker("wandb")
+        # pretty name for current step: step 50000 -> step 50k
+        cur_step_pretty = f"{int(step // 1000)}k" if step > 1000 else step
+        prefix_pretty = prefix.replace("/", "-")
+
+        # convert str data to a wandb compatible format
+        str_data = [[pred_descriptions[i], pred_prompts[i], transcriptions[i]] for i in range(len(pred_descriptions))]
+        # log as a table with the appropriate headers
+        wandb_tracker.log_table(
+            table_name=f"predictions/{prefix_pretty}-step-{cur_step_pretty}",
+            columns=["Target descriptions", "Target prompts", "Predicted transcriptions"],
+            data=str_data[:num_lines],
+            step=step,
+            commit=False,
+        )
+
+        # wandb can only loads 100 audios per step
+        wandb_tracker.log(
+            {
+                f"Speech samples/{prefix}": [
+                    Audio(
+                        audio,
+                        caption=f"{pred_prompts[i]} --- DESCRIPTION: {pred_descriptions[i]}",
+                        sample_rate=sampling_rate,
+                    )
+                    for (i, audio) in enumerate(audios[: min(len(audios), 100)])
+                ]
+            },
+            step=step,
+        )
+
+
+@dataclass
+class ModelArguments:
+    """
+    Arguments pertaining to which model/config/tokenizer we are going to fine-tune from.
+    """
+
+    model_name_or_path: str = field(
+        metadata={"help": "Path to pretrained model or model identifier from huggingface.co/models"}
+    )
+    config_name: Optional[str] = field(
+        default=None, metadata={"help": "Pretrained config name or path if not the same as model_name"}
+    )
+    feature_extractor_name: Optional[str] = field(
+        default=None, metadata={"help": "Pretrained feature extractor name or path if not the same as model_name"}
+    )
+    description_tokenizer_name: Optional[str] = field(
+        default=None, metadata={"help": "Pretrained description tokenizer name or path if not the same as model_name"}
+    )
+    prompt_tokenizer_name: Optional[str] = field(
+        default=None,
+        metadata={"help": "Pretrained prompt tokenizer name or path if not the same as description_tokenizer_name"},
+    )
+    cache_dir: Optional[str] = field(
+        default=None,
+        metadata={"help": "Where to store the pretrained models downloaded from huggingface.co"},
+    )
+    use_fast_tokenizer: bool = field(
+        default=True,
+        metadata={"help": "Whether to use one of the fast tokenizer (backed by the tokenizers library) or not."},
+    )
+    model_revision: str = field(
+        default="main",
+        metadata={"help": "The specific model version to use (can be a branch name, tag name or commit id)."},
+    )
+    pad_token_id: int = field(
+        default=None,
+        metadata={"help": "If specified, change the model pad token id."},
+    )
+    decoder_start_token_id: int = field(
+        default=None,
+        metadata={"help": "If specified, change the model decoder start token id."},
+    )
+    freeze_text_encoder: bool = field(
+        default=False,
+        metadata={"help": "Whether to freeze the text encoder."},
+    )
+    do_sample: bool = field(
+        default=True,
+        metadata={"help": "Whether to do sampling or greedy decoding."},
+    )
+    temperature: float = field(
+        default=1.0,
+        metadata={"help": "Temperature if sampling."},
+    )
+    max_length: int = field(
+        default=2580,
+        metadata={"help": "Generation max length."},
+    )
+    bandwidth: float = field(
+        default=6,
+        metadata={"help": "Audio encoder bandwidth."},
+    )
+    asr_model_name_or_path: str = field(
+        default="distil-whisper/distil-large-v2",
+        metadata={
+            "help": "Used to compute WER during evaluation. Path to pretrained model or model identifier from huggingface.co/models"
+        },
+    )
+    clap_model_name_or_path: str = field(
+        default="laion/larger_clap_music_and_speech",
+        metadata={
+            "help": "Used to compute audio similarity during evaluation. Path to pretrained model or model identifier from huggingface.co/models"
+        },
+    )
+
+
+@dataclass
+class DataTrainingArguments:
+    """
+    Arguments pertaining to what data we are going to input our model for training and eval.
+
+    Using `HfArgumentParser` we can turn this class
+    into argparse arguments to be able to specify them on
+    the command line.
+    """
+
+    train_dataset_name: str = field(
+        default=None,
+        metadata={
+            "help": "The name of the training dataset to use (via the datasets library). Load and combine "
+            "multiple datasets by separating dataset ids by a '+' symbol. For example, to load and combine "
+            " librispeech and common voice, set `train_dataset_name='librispeech_asr+common_voice'`."
+        },
+    )
+    train_dataset_config_name: Optional[str] = field(
+        default=None,
+        metadata={
+            "help": "The configuration name of the training dataset to use (via the datasets library). Load and combine "
+            "multiple datasets by separating dataset configs by a '+' symbol."
+        },
+    )
+    train_split_name: str = field(
+        default="train",
+        metadata={
+            "help": ("The name of the training data set split to use (via the datasets library). Defaults to 'train'")
+        },
+    )
+    train_dataset_samples: str = field(
+        default=None,
+        metadata={
+            "help": "Number of samples in the training data. Load and combine "
+            "multiple datasets by separating dataset samples by a '+' symbol."
+        },
+    )
+    train_metadata_dataset_name: str = field(
+        default=None,
+        metadata={
+            "help": "The name of the metadata training dataset to use (via the datasets library). Load and combine "
+            "multiple datasets by separating dataset ids by a '+' symbol. For example, to load and combine "
+            " librispeech and common voice, set `train_dataset_name='librispeech_asr+common_voice'`."
+        },
+    )
+    eval_dataset_name: str = field(
+        default=None,
+        metadata={
+            "help": "The name of the evaluation dataset to use (via the datasets library). Defaults to the training dataset name if unspecified."
+        },
+    )
+    eval_dataset_config_name: Optional[str] = field(
+        default=None,
+        metadata={
+            "help": "The configuration name of the evaluation dataset to use (via the datasets library). Defaults to the training dataset config name if unspecified"
+        },
+    )
+    eval_split_name: str = field(
+        default="test",
+        metadata={
+            "help": "The name of the evaluation data set split to use (via the datasets library). Defaults to 'test'"
+        },
+    )
+    eval_metadata_dataset_name: str = field(
+        default=None,
+        metadata={
+            "help": "The name of the metadata training dataset to use (via the datasets library). Load and combine "
+            "multiple datasets by separating dataset ids by a '+' symbol. For example, to load and combine "
+            " librispeech and common voice, set `train_dataset_name='librispeech_asr+common_voice'`."
+        },
+    )
+    target_audio_column_name: str = field(
+        default="audio",
+        metadata={"help": "The name of the dataset column containing the target audio data. Defaults to 'audio'"},
+    )
+    description_column_name: str = field(
+        default=None,
+        metadata={"help": "The name of the dataset column containing the description text data. Defaults to 'None'."},
+    )
+    prompt_column_name: str = field(
+        default=None,
+        metadata={"help": "The name of the dataset column containing the prompt text data. Defaults to 'None'."},
+    )
+    overwrite_cache: bool = field(
+        default=False, metadata={"help": "Overwrite the cached preprocessed datasets or not."}
+    )
+    preprocessing_num_workers: Optional[int] = field(
+        default=None,
+        metadata={"help": "The number of processes to use for the preprocessing."},
+    )
+    max_train_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": (
+                "For debugging purposes or quicker training, truncate the number of training examples to this "
+                "value if set."
+            )
+        },
+    )
+    max_eval_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": (
+                "For debugging purposes or quicker training, truncate the number of validation examples to this "
+                "value if set."
+            )
+        },
+    )
+    max_duration_in_seconds: float = field(
+        default=35.0,
+        metadata={
+            "help": (
+                "Filter audio files that are longer than `max_duration_in_seconds` seconds to 'max_duration_in_seconds`."
+                "Also, used to set maximum audio length if `pad_to_max_length=True`."
+            )
+        },
+    )
+    min_duration_in_seconds: float = field(
+        default=0.0, metadata={"help": "Filter audio files that are shorter than `min_duration_in_seconds` seconds"}
+    )
+    max_text_length: int = field(
+        default=500, metadata={"help": "If set, max description lengths in number of characters."}
+    )
+    max_prompt_token_length: int = field(
+        default=None,
+        metadata={
+            "help": (
+                "If set, filter samples with prompts that are longer than `max_prompt_token_length` tokens."
+                "Also, used to set maximum prompt token length if `pad_to_max_length=True`."
+            )
+        },
+    )
+    max_description_token_length: int = field(
+        default=None,
+        metadata={
+            "help": (
+                "If set, filter samples with descriptions that are longer than `max_description_token_length` tokens."
+                "Also, used to set maximum desription token length if `pad_to_max_length=True`."
+            )
+        },
+    )
+    pad_to_max_length: bool = field(
+        default=False,
+        metadata={
+            "help": (
+                "If `True`, pad audio, prompt and description to a maximum length set with respectively "
+                "`max_duration_in_seconds`, `max_prompt_token_length`, `max_description_token_length`."
+            )
+        },
+    )
+    preprocessing_only: bool = field(
+        default=False,
+        metadata={
+            "help": (
+                "Whether to only do data preprocessing and skip training. This is especially useful when data"
+                " preprocessing errors out in distributed training due to timeout. In this case, one should run the"
+                " preprocessing in a non-distributed setup with `preprocessing_only=True` so that the cached datasets"
+                " can consequently be loaded in distributed training."
+                " In this training script, `save_to_disk` must be set to the path in which the dataset should be saved. "
+            )
+        },
+    )
+    token: str = field(
+        default=None,
+        metadata={
+            "help": (
+                "The token to use as HTTP bearer authorization for remote files. If not specified, will use the token "
+                "generated when running `huggingface-cli login` (stored in `~/.huggingface`)."
+            )
+        },
+    )
+    use_auth_token: bool = field(
+        default=None,
+        metadata={
+            "help": "The `use_auth_token` argument is deprecated and will be removed in v4.34. Please use `token` instead."
+        },
+    )
+    trust_remote_code: bool = field(
+        default=False,
+        metadata={
+            "help": (
+                "Whether or not to allow for custom models defined on the Hub in their own modeling files. This option "
+                "should only be set to `True` for repositories you trust and in which you have read the code, as it will "
+                "execute code present on the Hub on your local machine."
+            )
+        },
+    )
+    add_audio_samples_to_wandb: bool = field(
+        default=False,
+        metadata={"help": "If set and if `wandb` in args.report_to, will add generated audio samples to wandb logs."},
+    )
+    id_column_name: str = field(default=None, metadata={"help": "id column name."})
+    wandb_project: str = field(
+        default="parler-speech",
+        metadata={"help": "The name of the wandb project."},
+    )
+    save_to_disk: str = field(
+        default=None,
+        metadata={
+            "help": "If set, will save the dataset to this path if this is an empyt folder. If not empty, will load the datasets from it."
+        },
+    )
+    temporary_save_to_disk: str = field(default=None, metadata={"help": "Temporarily save audio labels here."})
+    pad_to_multiple_of: Optional[int] = field(
+        default=2,
+        metadata={"help": ("Pad to multiple of for tokenizers.")},
+    )
+
+
+@dataclass
+class ParlerTTSTrainingArguments(Seq2SeqTrainingArguments):
+    dtype: Optional[str] = field(
+        default="float32",
+        metadata={
+            "help": (
+                "The data type (dtype) in which to run training. One of `float32` (full-precision), "
+                "`float16` or `bfloat16` (both half-precision)."
+            )
+        },
+    )
+    audio_encoder_per_device_batch_size: int = field(
+        default=8,
+        metadata={"help": ("Specify the batch size of the audio encoding pre-processing steps.")},
+    )
+
+
+@dataclass
+class DataCollatorEncodecWithPadding:
+    """
+    Data collator that will dynamically pad the inputs received to the longest sequence in the batch or
+    to `max_length` if `max_length` is set and `padding=max_length`.
+    """
+
+    feature_extractor: AutoFeatureExtractor
+    audio_column_name: str
+    feature_extractor_input_name: Optional[str] = "input_values"
+    max_length: Optional[int] = None
+    padding: Optional[str] = "longest"
+
+    def __call__(self, features: List[Dict[str, Union[List[int], torch.Tensor]]]) -> Dict[str, torch.Tensor]:
+        # split inputs and labels since they have to be of different lengths and need
+        # different padding methods
+        audios = [feature[self.audio_column_name]["array"] for feature in features]
+        len_audio = [len(audio) for audio in audios]
+
+        batch = self.feature_extractor(audios, return_tensors="pt", padding=self.padding, max_length=self.max_length, sampling_rate=self.feature_extractor.sampling_rate)
+        batch["len_audio"] = torch.tensor(len_audio).unsqueeze(1)
+        return batch
+
+
+@dataclass
+class DataCollatorParlerTTSWithPadding:
+    """
+    Data collator that will dynamically pad the inputs received.
+    Args:
+        prompt_tokenizer (:class:`~transformers.AutoTokenizer`)
+            The prompt_tokenizer used for proccessing the data.
+        description_tokenizer (:class:`~transformers.AutoTokenizer`)
+            The description_tokenizer used for proccessing the data.
+        padding (:obj:`bool`, :obj:`str` or :class:`~transformers.tokenization_utils_base.PaddingStrategy`, `optional`, defaults to :obj:`True`):
+            Select a strategy to pad the returned sequences (according to the model's padding side and padding index)
+            among:
+            * :obj:`True` or :obj:`'longest'`: Pad to the longest sequence in the batch (or no padding if only a single
+              sequence if provided).
+            * :obj:`'max_length'`: Pad to a maximum length specified with the argument :obj:`max_length` or to the
+              maximum acceptable input length for the model if that argument is not provided.
+            * :obj:`False` or :obj:`'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of
+              different lengths).
+        pad_to_multiple_of (:obj:`int`, `optional`):
+            If set will pad the sequence to a multiple of the provided value.
+            This is especially useful to enable the use of Tensor Cores on NVIDIA hardware with compute capability >=
+            7.5 (Volta).
+    """
+
+    prompt_tokenizer: AutoTokenizer
+    description_tokenizer: AutoTokenizer
+    padding: Union[bool, str] = "longest"
+    pad_to_multiple_of: Optional[int] = None
+    prompt_max_length: Optional[int] = None
+    description_max_length: Optional[int] = None
+    audio_max_length: Optional[int] = None
+
+    def __call__(self, features: List[Dict[str, Union[List[int], torch.Tensor]]]) -> Dict[str, torch.Tensor]:
+        # split inputs and labels since they have to be of different lengths and need
+        # different padding methods
+
+        labels = [torch.tensor(feature["labels"]).transpose(0, 1) for feature in features]
+        # (bsz, seq_len, num_codebooks)
+        labels = torch.nn.utils.rnn.pad_sequence(labels, batch_first=True, padding_value=-100)
+        if self.audio_max_length is not None and self.padding == "max_length":
+            labels = torch.nn.functional.pad(labels, pad=(0, 0, 0, max(self.audio_max_length - labels.shape[1], 0)))
+
+        input_ids = [{"input_ids": feature["input_ids"]} for feature in features]
+
+        input_ids = self.description_tokenizer.pad(
+            input_ids,
+            return_tensors="pt",
+            padding=self.padding,
+            pad_to_multiple_of=self.pad_to_multiple_of,
+            max_length=self.description_max_length,
+        )
+
+        batch = {"labels": labels, **input_ids}
+
+        if self.audio_max_length is not None and self.padding == "max_length":
+            # if we do torch.compile, we need to also specify the attention_mask
+            decoder_attention_mask = torch.ones(labels.shape[:2], dtype=input_ids["attention_mask"].dtype)
+            batch["decoder_attention_mask"] = decoder_attention_mask
+
+        prompt_input_ids = [{"input_ids": feature["prompt_input_ids"]} for feature in features]
+        prompt_input_ids = self.prompt_tokenizer.pad(
+            prompt_input_ids,
+            return_tensors="pt",
+            padding=self.padding,
+            pad_to_multiple_of=self.pad_to_multiple_of,
+            max_length=self.prompt_max_length,
+        )
+
+        batch["prompt_input_ids"] = prompt_input_ids["input_ids"]
+        if "attention_mask" in prompt_input_ids:
+            batch["prompt_attention_mask"] = prompt_input_ids["attention_mask"]
+
+        return batch
+
+
+def convert_dataset_str_to_list(
+    dataset_names,
+    dataset_config_names,
+    metadata_dataset_names=None,
+    splits=None,
+    dataset_samples=None,
+    default_split="train",
+):
+    if isinstance(dataset_names, str):
+        dataset_names = dataset_names.split("+")
+        dataset_config_names = dataset_config_names.split("+")
+        splits = splits.split("+") if splits is not None else None
+        dataset_samples = dataset_samples.split("+") if dataset_samples is not None else None
+        metadata_dataset_names = metadata_dataset_names.split("+") if metadata_dataset_names is not None else None
+
+    # basic checks to ensure we've got the right number of datasets/configs/splits/columns/probs
+    if len(dataset_names) != len(dataset_config_names):
+        raise ValueError(
+            f"Ensure one config is passed for each dataset, got {len(dataset_names)} datasets and"
+            f" {len(dataset_config_names)} configs."
+        )
+
+    if splits is not None and len(splits) != len(dataset_names):
+        raise ValueError(
+            f"Ensure one split is passed for each dataset, got {len(dataset_names)} datasets and {len(splits)} splits."
+        )
+
+    if metadata_dataset_names is not None and len(metadata_dataset_names) != len(dataset_names):
+        raise ValueError(
+            f"Ensure one metadata dataset is passed for each dataset, got {len(dataset_names)} datasets and {len(metadata_dataset_names)} metadata datasets."
+        )
+
+    if dataset_samples is not None:
+        if len(dataset_samples) != len(dataset_names):
+            raise ValueError(
+                f"Ensure one sample is passed for each dataset, got {len(dataset_names)} datasets and "
+                f"{len(dataset_samples)} samples."
+            )
+        dataset_samples = [float(ds_sample) for ds_sample in dataset_samples]
+    else:
+        dataset_samples = [None] * len(dataset_names)
+
+    splits = splits if splits is not None else [default_split for _ in range(len(dataset_names))]
+
+    dataset_names_dict = []
+    for i, ds_name in enumerate(dataset_names):
+        dataset_names_dict.append(
+            {
+                "name": ds_name,
+                "config": dataset_config_names[i],
+                "split": splits[i],
+                "metadata_dataset_name": metadata_dataset_names[i],
+                "samples": dataset_samples[i],
+            }
+        )
+    return dataset_names_dict
+
+
+def load_multiple_datasets(
+    accelerator: Accelerator,
+    dataset_names: Union[List, str],
+    dataset_config_names: Union[List, str],
+    metadata_dataset_names: Optional[str] = None,
+    splits: Optional[Union[List, str]] = None,
+    label_column_names: Optional[List] = None,
+    stopping_strategy: Optional[str] = "first_exhausted",
+    dataset_samples: Optional[Union[List, np.array]] = None,
+    streaming: Optional[bool] = False,
+    seed: Optional[int] = None,
+    id_column_name: Optional[str] = None,
+    columns_to_keep: Optional[Set[str]] = None,
+    prompt_column_name: Optional[str] = None,
+    sampling_rate: Optional[int] = None,
+    audio_column_name: Optional[str] = None,
+    **kwargs,
+) -> Union[Dataset, IterableDataset]:
+    dataset_names_dict = convert_dataset_str_to_list(
+        dataset_names, dataset_config_names, metadata_dataset_names, splits, label_column_names, dataset_samples
+    )
+
+    if dataset_samples is not None:
+        dataset_samples = [ds_dict["samples"] for ds_dict in dataset_names_dict]
+        probabilities = np.array(dataset_samples) / np.sum(dataset_samples)
+    else:
+        probabilities = None
+
+    all_datasets = []
+    # iterate over the datasets we want to interleave
+    for dataset_dict in tqdm(dataset_names_dict, desc="Combining datasets..."):
+        with accelerator.main_process_first():
+            dataset = load_dataset(
+                dataset_dict["name"],
+                dataset_dict["config"],
+                split=dataset_dict["split"],
+                streaming=streaming,
+                **kwargs,
+            )
+            dataset_features = dataset.features.keys()
+
+            if sampling_rate is not None and audio_column_name is not None:
+                # resample target audio
+                dataset = dataset.cast_column(audio_column_name, datasets.features.Audio(sampling_rate=sampling_rate))
+
+            metadata_dataset_name = dataset_dict["metadata_dataset_name"]
+            if metadata_dataset_name is not None:
+                logger.info(
+                    f'Merging {dataset_dict["name"]} - {dataset_dict["split"]} with {metadata_dataset_name} - {dataset_dict["split"]}'
+                )
+                metadata_dataset = load_dataset(
+                    metadata_dataset_name,
+                    dataset_dict["config"],
+                    split=dataset_dict["split"],
+                    streaming=streaming,
+                    **kwargs,
+                )
+
+                # TODO(YL): I forgot to create unique ids for MLS english.
+                # To iterate faster, I bypass the original id check and do another one. - Done once because assuming it won't change next time
+                # if dataset_dict["name"] == "parler-tts/mls_eng_10k":
+                #     def concat_ids(book_id, speaker_id, begin_time):
+                #         return {"id": f"{book_id}_{speaker_id}_{str(begin_time).replace('.', '_')}"}
+                #     dataset = dataset.map(concat_ids, input_columns=["book_id", "speaker_id", "begin_time"], num_proc=24)
+                #     metadata_dataset = metadata_dataset.map(concat_ids, input_columns=["book_id", "speaker_id", "begin_time"], num_proc=24)
+                #     metadata_dataset = metadata_dataset.rename_column(id_column_name, f"metadata_{id_column_name}")
+
+                if dataset_dict["name"] != "parler-tts/mls_eng_10k":
+                    if id_column_name is not None and id_column_name not in dataset.column_names:
+                        raise ValueError(
+                            f"id_column_name={id_column_name} but has not been found in the dataset columns"
+                            f"- one of {', '.join(list(dataset.column_names))}."
+                        )
+                    if id_column_name is not None and id_column_name not in metadata_dataset.column_names:
+                        raise ValueError(
+                            f"id_column_name={id_column_name} but has not been found in the metadata dataset columns"
+                            f"- one of {', '.join(list(metadata_dataset.column_names))}."
+                        )
+                    elif id_column_name is not None:
+                        metadata_dataset = metadata_dataset.rename_column(id_column_name, f"metadata_{id_column_name}")
+
+                metadata_columns_to_remove = set(metadata_dataset.column_names).intersection(set(dataset.column_names))
+
+                if prompt_column_name is not None:
+                    # We might have applied some transformations to the prompts (e.g  punctuation restoration)
+                    # so we make sure to remove it from the original dataset
+                    if prompt_column_name in dataset.column_names:
+                        logger.info(
+                            f"REMOVE {prompt_column_name} from dataset {dataset_dict['name']} - dataset_dict['split']"
+                        )
+                        dataset.remove_columns(prompt_column_name)
+
+                metadata_columns_to_remove = set(metadata_dataset.column_names).intersection(set(dataset.column_names))
+                metadata_dataset = metadata_dataset.remove_columns(metadata_columns_to_remove)
+
+                dataset = concatenate_datasets([dataset, metadata_dataset], axis=1)
+
+                if id_column_name is not None and dataset_dict["name"] != "parler-tts/mls_eng_10k":
+                    if (
+                        len(
+                            dataset.filter(
+                                lambda id1, id2: id1 != id2,
+                                input_columns=[id_column_name, f"metadata_{id_column_name}"],
+                            )
+                        )
+                        != 0
+                    ):
+                        raise ValueError(
+                            f"Concatenate didn't work. Some ids don't correspond on dataset {dataset_dict['name']}"
+                        )
+
+                dataset_features = dataset.features.keys()
+
+            if columns_to_keep is not None:
+                dataset = dataset.remove_columns(set(dataset_features - columns_to_keep))
+        all_datasets.append(dataset)
+
+    if len(all_datasets) == 1:
+        # we have a single dataset so just return it as is
+        return all_datasets[0]
+
+    if streaming:
+        interleaved_dataset = interleave_datasets(
+            all_datasets,
+            stopping_strategy=stopping_strategy,
+            probabilities=probabilities,
+            seed=seed,
+        )
+    else:
+        with accelerator.main_process_first():
+            interleaved_dataset = concatenate_datasets(all_datasets)
+
+    return interleaved_dataset
+
+
+def main():
+    # See all possible arguments in src/transformers/training_args.py
+    # or by passing the --help flag to this script.
+    # We now keep distinct sets of args, for a cleaner separation of concerns.
+
+    parser = HfArgumentParser((ModelArguments, DataTrainingArguments, ParlerTTSTrainingArguments))
+    if len(sys.argv) == 2 and sys.argv[1].endswith(".json"):
+        # If we pass only one argument to the script and it's the path to a json file,
+        # let's parse it to get our arguments.
+        model_args, data_args, training_args = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1]))
+    else:
+        model_args, data_args, training_args = parser.parse_args_into_dataclasses()
+
+    # Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The
+    # information sent is the one passed as arguments along with your Python/PyTorch versions.
+    send_example_telemetry("run_parler_tts", model_args, data_args)
+
+    if training_args.dtype == "float16":
+        mixed_precision = "fp16"
+    elif training_args.dtype == "bfloat16":
+        mixed_precision = "bf16"
+    else:
+        mixed_precision = "no"
+
+    if data_args.pad_to_max_length and (
+        data_args.max_duration_in_seconds is None
+        or data_args.max_prompt_token_length is None
+        or data_args.max_description_token_length is None
+    ):
+        raise ValueError(
+            "`pad_to_max_length` is `True` but one of the following parameters has not been set: `max_duration_in_seconds`, `max_prompt_token_length`, `max_description_token_length`"
+        )
+
+    padding = "max_length" if data_args.pad_to_max_length else "longest"
+
+    ####### A. Preparation
+    kwargs_handlers = [InitProcessGroupKwargs(timeout=timedelta(minutes=60))]
+    if training_args.torch_compile:
+        # TODO(YL): add more compile modes?
+        kwargs_handlers.append(TorchDynamoPlugin(backend="inductor", mode="default"))  # reduce-overhead
+
+    accelerator = Accelerator(
+        gradient_accumulation_steps=training_args.gradient_accumulation_steps,
+        mixed_precision=mixed_precision,
+        log_with=training_args.report_to,
+        project_dir=training_args.output_dir,
+        kwargs_handlers=kwargs_handlers,
+    )
+
+    accelerator.init_trackers(
+        project_name=data_args.wandb_project,
+        config={
+            "learning_rate": training_args.learning_rate,
+            "model_name_or_path": model_args.model_name_or_path,
+            "num_train_epochs": training_args.num_train_epochs,
+            "gradient_accumulation_steps": training_args.gradient_accumulation_steps,
+            "per_device_train_batch_size": training_args.per_device_train_batch_size,
+            "global_batch_size": training_args.per_device_train_batch_size * accelerator.num_processes,
+            "mixed_precision": mixed_precision,
+            "lr_scheduler_type": training_args.lr_scheduler_type,
+            "warmup_steps": training_args.warmup_steps,
+            "freeze_text_encoder": model_args.freeze_text_encoder,
+            "max_duration_in_seconds": data_args.max_duration_in_seconds,
+            "weight_decay": training_args.weight_decay,
+            "adam_beta1": training_args.adam_beta1,
+            "adam_beta2": training_args.adam_beta2,
+            "temperature": model_args.temperature,
+        },
+    )
+
+    # Detecting last checkpoint and eventually continue from last checkpoint
+    last_checkpoint = None
+    if os.path.isdir(training_args.output_dir) and training_args.do_train and not training_args.overwrite_output_dir:
+        last_checkpoint = get_last_checkpoint(training_args.output_dir)
+        if last_checkpoint is None and len(os.listdir(training_args.output_dir)) > 0:
+            raise ValueError(
+                f"Output directory ({training_args.output_dir}) already exists and is not empty. "
+                "Use --overwrite_output_dir to overcome."
+            )
+        elif last_checkpoint is not None and training_args.resume_from_checkpoint is None:
+            logger.info(
+                f"Checkpoint detected, resuming training at {last_checkpoint}. To avoid this behavior, change "
+                "the `--output_dir` or add `--overwrite_output_dir` to train from scratch."
+            )
+
+    # Setup logging
+    logging.basicConfig(
+        format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
+        datefmt="%m/%d/%Y %H:%M:%S",
+        handlers=[logging.StreamHandler(sys.stdout)],
+    )
+    logger.setLevel(logging.INFO if accelerator.is_main_process else logging.WARN)
+
+    # Log a small summary on each proces
+    logger.warning(
+        f"Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}, "
+        f"distributed training: {training_args.parallel_mode.value == 'distributed'}, 16-bits training: {training_args.fp16}"
+    )
+
+    # Set the verbosity to info of the Transformers logger (on main process only)
+    if accelerator.is_local_main_process:
+        datasets.utils.logging.set_verbosity_warning()
+        transformers.utils.logging.set_verbosity_info()
+    else:
+        datasets.utils.logging.set_verbosity_error()
+        transformers.utils.logging.set_verbosity_error()
+
+    logger.info("Training/evaluation parameters %s", training_args)
+
+    # Set seed before initializing model.
+    set_seed(training_args.seed)
+    num_workers = data_args.preprocessing_num_workers
+
+    # 1. First, lett's instantiate the feature extractor, tokenizers and model
+    # Note for distributed training, the .from_pretrained methods guarantee that only
+    # one local process can concurrently download model & vocab.
+
+    # load feature extractor
+    feature_extractor = AutoFeatureExtractor.from_pretrained(
+        model_args.feature_extractor_name or model_args.model_name_or_path,
+        cache_dir=model_args.cache_dir,
+        token=data_args.token,
+        trust_remote_code=data_args.trust_remote_code,
+    )
+    sampling_rate = feature_extractor.sampling_rate
+
+    # load prompt tokenizer
+    prompt_tokenizer = AutoTokenizer.from_pretrained(
+        model_args.prompt_tokenizer_name or model_args.description_tokenizer_name or model_args.model_name_or_path,
+        cache_dir=model_args.cache_dir,
+        token=data_args.token,
+        trust_remote_code=data_args.trust_remote_code,
+        use_fast=model_args.use_fast_tokenizer,
+        padding_side="left",  # prompt has to be padded on the left bc it's preprend to codebooks hidden states
+    )
+
+    # load description tokenizer
+    description_tokenizer = AutoTokenizer.from_pretrained(
+        model_args.description_tokenizer_name or model_args.model_name_or_path,
+        cache_dir=model_args.cache_dir,
+        token=data_args.token,
+        trust_remote_code=data_args.trust_remote_code,
+        use_fast=model_args.use_fast_tokenizer,
+    )
+
+    if model_args.use_fast_tokenizer:
+        logger.warning(
+            "Disabling fast tokenizer warning: https://github.com/huggingface/transformers/blob/main/src/transformers/tokenization_utils_base.py#L3231-L3235"
+        )
+        prompt_tokenizer.deprecation_warnings["Asking-to-pad-a-fast-tokenizer"] = True
+        description_tokenizer.deprecation_warnings["Asking-to-pad-a-fast-tokenizer"] = True
+
+    # 2. Now, let's load the dataset
+
+    if data_args.save_to_disk is not None:
+        os.makedirs(data_args.save_to_disk, exist_ok=True)
+
+    # assume that the dataset has been saved to `save_to_disk` if the latter is not empty
+    dataset_was_precomputed = len(os.listdir(data_args.save_to_disk)) > 0
+    if dataset_was_precomputed:
+        vectorized_datasets = datasets.load_from_disk(data_args.save_to_disk)
+    else:
+        raw_datasets = DatasetDict()
+
+        columns_to_keep = {
+            "target_audio_column_name": data_args.target_audio_column_name,
+            "prompt_column_name": data_args.prompt_column_name,
+        }
+        if data_args.description_column_name is not None:
+            columns_to_keep["description_column_name"] = data_args.description_column_name
+
+        if training_args.do_train:
+            raw_datasets["train"] = load_multiple_datasets(
+                accelerator,
+                data_args.train_dataset_name,
+                data_args.train_dataset_config_name,
+                metadata_dataset_names=data_args.train_metadata_dataset_name,
+                splits=data_args.train_split_name,
+                dataset_samples=data_args.train_dataset_samples,
+                seed=training_args.seed,
+                cache_dir=model_args.cache_dir,
+                num_proc=data_args.preprocessing_num_workers,
+                id_column_name=data_args.id_column_name,
+                columns_to_keep=columns_to_keep.values(),
+                prompt_column_name=data_args.prompt_column_name,
+                audio_column_name=data_args.target_audio_column_name,
+                sampling_rate=sampling_rate,
+                # streaming=data_args.streaming, TODO(SG): optionally enable streaming mode
+            )
+
+            for key in columns_to_keep:
+                if columns_to_keep[key] not in raw_datasets["train"].column_names:
+                    raise ValueError(
+                        f"--{key} '{columns_to_keep[key]}' not found in dataset '{data_args.train_dataset_name}'."
+                        f" Make sure to set `--{key}` to the correct audio column - one of"
+                        f" {', '.join(raw_datasets['train'].column_names)}."
+                    )
+
+            if data_args.max_train_samples is not None:
+                raw_datasets["train"] = raw_datasets["train"].select(range(data_args.max_train_samples))
+
+        if training_args.do_eval:
+            raw_datasets["eval"] = load_multiple_datasets(
+                accelerator,
+                data_args.eval_dataset_name if data_args.eval_dataset_name else data_args.train_dataset_name,
+                data_args.eval_dataset_config_name
+                if data_args.eval_dataset_config_name
+                else data_args.train_dataset_config_name,
+                metadata_dataset_names=data_args.eval_metadata_dataset_name,
+                splits=data_args.eval_split_name,
+                cache_dir=model_args.cache_dir,
+                num_proc=data_args.preprocessing_num_workers,
+                id_column_name=data_args.id_column_name,
+                columns_to_keep=columns_to_keep.values(),
+                prompt_column_name=data_args.prompt_column_name,
+                audio_column_name=data_args.target_audio_column_name,
+                sampling_rate=sampling_rate,
+                # streaming=data_args.streaming, TODO(SG): optionally enable streaming mode
+            )
+
+            if data_args.max_eval_samples is not None:
+                raw_datasets["eval"] = (
+                    raw_datasets["eval"].shuffle(seed=training_args.seed).select(range(data_args.max_eval_samples))
+                )
+
+    # 3. Next, let's load the config.
+    config = ParlerTTSConfig.from_pretrained(
+        model_args.model_name_or_path,
+        cache_dir=model_args.cache_dir,
+        token=data_args.token,
+        trust_remote_code=data_args.trust_remote_code,
+    )
+
+    # update pad token id and decoder_start_token_id
+    config.update(
+        {
+            "pad_token_id": model_args.pad_token_id if model_args.pad_token_id is not None else config.pad_token_id,
+            "decoder_start_token_id": (
+                model_args.decoder_start_token_id
+                if model_args.decoder_start_token_id is not None
+                else config.decoder_start_token_id
+            ),
+        }
+    )
+
+    # create model
+    model = ParlerTTSForConditionalGeneration.from_pretrained(
+        model_args.model_name_or_path,
+        cache_dir=model_args.cache_dir,
+        config=config,
+        token=data_args.token,
+        trust_remote_code=data_args.trust_remote_code,
+    )
+
+    # enable gradient checkpointing if necessary
+    if training_args.gradient_checkpointing:
+        model.gradient_checkpointing_enable()
+
+    # 4. Now we preprocess the datasets including loading the audio, resampling and normalization
+    # Thankfully, `datasets` takes care of automatically loading and resampling the audio,
+    # so that we just need to set the correct target sampling rate and normalize the input
+    # via the `feature_extractor`
+
+    # derive max & min input length for sample rate & max duration
+    sampling_rate = feature_extractor.sampling_rate
+    max_target_length = data_args.max_duration_in_seconds * sampling_rate
+    min_target_length = data_args.min_duration_in_seconds * sampling_rate
+    target_audio_column_name = data_args.target_audio_column_name
+    description_column_name = data_args.description_column_name
+    prompt_column_name = data_args.prompt_column_name
+    feature_extractor_input_name = feature_extractor.model_input_names[0]
+    audio_encoder_pad_token_id = config.decoder.pad_token_id
+    audio_encoder_eos_token_id = config.decoder.eos_token_id
+    audio_encoder_bos_token_id = model.generation_config.decoder_start_token_id
+    max_length = model.generation_config.max_length
+    num_codebooks = model.decoder.config.num_codebooks
+    bandwidth = model_args.bandwidth
+
+    # Freeze Encoders
+    model.freeze_encoders(model_args.freeze_text_encoder)
+
+    # Test all gather - used for warmout and avoiding timeout
+    test_tensor = torch.tensor([accelerator.process_index], device=accelerator.device)
+    gathered_tensor = accelerator.gather(test_tensor)
+    print("gathered_tensor", gathered_tensor)
+    accelerator.wait_for_everyone()
+
+    if not dataset_was_precomputed:
+        # Filter on text length
+        if description_column_name is not None and data_args.max_text_length is not None:
+            with accelerator.main_process_first():
+                # filter description that is shorter than max_text_length
+                raw_datasets = raw_datasets.filter(
+                    lambda x: len(x) < data_args.max_text_length,
+                    num_proc=num_workers,
+                    input_columns=[description_column_name],
+                )
+
+        # Preprocessing the dataset.
+        # We need to tokenize the texts.
+        def pass_through_processors(description, prompt):
+            batch = {}
+
+            batch["input_ids"] = description_tokenizer(description.strip())["input_ids"]
+            batch["prompt_input_ids"] = prompt_tokenizer(prompt.strip())["input_ids"]
+
+            return batch
+
+        with accelerator.main_process_first():
+            # this is a trick to avoid to rewrite the entire audio column which takes ages
+            vectorized_datasets = raw_datasets.map(
+                pass_through_processors,
+                remove_columns=next(iter(raw_datasets.values())).column_names,
+                input_columns=[description_column_name, prompt_column_name],
+                num_proc=num_workers,
+                desc="preprocess datasets",
+            )
+
+        # We use Accelerate to perform distributed inference
+        # T5 doesn't support fp16
+        autocast_kwargs = AutocastKwargs(enabled=(mixed_precision != "fp16"))
+
+        # Now we encode the audio labels with encodec.
+        ####### B. Encode audio
+
+        logger.info("*** Encode target audio with encodec ***")
+
+        # no need to prepare audio_decoder because used for inference without mixed precision
+        # see: https://huggingface.co/docs/accelerate/main/en/package_reference/accelerator#accelerate.Accelerator.prepare
+        if training_args.torch_compile:
+            audio_decoder = accelerator.prepare_model(model.audio_encoder, evaluation_mode=True)
+        else:
+            audio_decoder = model.audio_encoder
+
+        encoder_data_collator = DataCollatorEncodecWithPadding(
+            feature_extractor,
+            audio_column_name=target_audio_column_name,
+            feature_extractor_input_name=feature_extractor_input_name,
+            max_length=max_target_length,
+            padding=padding,
+        )
+
+        def apply_audio_decoder(batch):
+            len_audio = batch.pop("len_audio")
+            audio_decoder.to(batch["input_values"].device).eval()
+            with torch.no_grad():
+                labels = audio_decoder.encode(**batch, bandwidth=bandwidth)["audio_codes"]
+            output = {}
+            output["len_audio"] = len_audio
+            # (1, bsz, codebooks, seq_len) -> (bsz, seq_len, codebooks)
+            output["labels"] = labels.squeeze(0).transpose(1, 2)
+            output["ratio"] = torch.ones_like(len_audio) * labels.shape[-1] / len_audio.max()
+            return output
+
+        for split in vectorized_datasets:
+            data_loader = DataLoader(
+                raw_datasets[split],
+                batch_size=training_args.audio_encoder_per_device_batch_size,
+                collate_fn=encoder_data_collator,
+                num_workers=training_args.dataloader_num_workers,
+                pin_memory=True,
+            )
+            data_loader = accelerator.prepare(data_loader)
+
+            all_generated_labels = []
+            all_lens = []
+            for batch in tqdm(data_loader, disable=not accelerator.is_local_main_process):
+                generate_labels = apply_audio_decoder(batch)
+                generate_labels = accelerator.pad_across_processes(generate_labels, dim=1, pad_index=0)
+                generate_labels = accelerator.gather_for_metrics(generate_labels)
+
+                if accelerator.is_main_process:
+                    lab = generate_labels["labels"].cpu().transpose(1, 2).to(torch.int16)
+                    rat = generate_labels["ratio"].cpu().squeeze()
+                    lens = generate_labels["len_audio"].cpu().squeeze()
+                    lab = [l[:, : int(ratio * length)] for (l, ratio, length) in zip(lab, rat, lens)]
+
+                    all_generated_labels.extend(lab)
+                    all_lens.extend(lens)
+
+            # (1, codebooks, seq_len) where seq_len=1
+            bos_labels = torch.ones((1, num_codebooks, 1)) * audio_encoder_bos_token_id
+
+            if accelerator.is_main_process:
+                tmp_labels = Dataset.from_dict({"labels": all_generated_labels, "target_length": all_lens})
+                tmp_labels.save_to_disk(
+                    os.path.join(data_args.temporary_save_to_disk, split),
+                    num_proc=1 if split == "eval" else data_args.preprocessing_num_workers,
+                )
+            accelerator.wait_for_everyone()
+            del all_generated_labels
+
+            tmp_labels = datasets.load_from_disk(os.path.join(data_args.temporary_save_to_disk, split))
+            with accelerator.main_process_first():
+                vectorized_datasets[split] = concatenate_datasets([vectorized_datasets[split], tmp_labels], axis=1)
+
+            def postprocess_dataset(labels):
+                # (1, codebooks, seq_len)
+                labels = torch.tensor(labels).unsqueeze(0)
+                # add bos
+                labels = torch.cat([bos_labels, labels], dim=-1)
+
+                labels, delay_pattern_mask = build_delay_pattern_mask(
+                    labels,
+                    bos_token_id=audio_encoder_bos_token_id,
+                    pad_token_id=audio_encoder_eos_token_id,
+                    max_length=labels.shape[-1] + num_codebooks,
+                    num_codebooks=num_codebooks,
+                )
+
+                # the first ids of the delay pattern mask are precisely labels, we use the rest of the labels mask
+                # to take care of EOS
+                # we want labels to look like this:
+                #  - [B, a, b, E, E, E, E]
+                #  - [B, B, c, d, E, E, E]
+                #  - [B, B, B, e, f, E, E]
+                #  - [B, B, B, B, g, h, E]
+                labels = torch.where(delay_pattern_mask == -1, audio_encoder_eos_token_id, delay_pattern_mask)
+
+                # the first timestamp is associated to a row full of BOS, let's get rid of it
+                # we also remove the last timestampts (full of PAD)
+                output = {"labels": labels[:, 1:]}
+                return output
+
+            with accelerator.main_process_first():
+                vectorized_datasets[split] = vectorized_datasets[split].map(
+                    postprocess_dataset,
+                    num_proc=data_args.preprocessing_num_workers,  # this one is resource consuming if many processor.
+                    input_columns=["labels"],
+                    desc="Postprocessing labeling",
+                )
+
+        accelerator.free_memory()
+        del generate_labels, all_lens
+
+        with accelerator.main_process_first():
+            # NOTE: filtering is done at the end because in the `datasets` library, caching audio files is done after most operations
+            # caching audio files is time and disk-space consuming, so we want to avoid it at all costs, especially for large (>1Kh) audio datasets.
+            # That's also why we avoid to concat the processed datasets (vectorized_datasets) with the audio column present in raw_datasets.
+
+            def is_audio_in_length_range(length):
+                return length > min_target_length and length < max_target_length
+
+            # filter data that is shorter than min_target_length
+            vectorized_datasets = vectorized_datasets.filter(
+                is_audio_in_length_range,
+                num_proc=num_workers,
+                input_columns=["target_length"],
+            )
+
+            if description_column_name is not None and data_args.max_description_token_length is not None:
+                with accelerator.main_process_first():
+                    # filter description that is shorter than max_text_length
+                    vectorized_datasets = vectorized_datasets.filter(
+                        lambda x: len(x) < data_args.max_description_token_length,
+                        num_proc=num_workers,
+                        input_columns=["input_ids"],
+                    )
+
+            if data_args.max_prompt_token_length is not None:
+                with accelerator.main_process_first():
+                    # filter description that is shorter than max_text_length
+                    vectorized_datasets = vectorized_datasets.filter(
+                        lambda x: len(x) < data_args.max_prompt_token_length,
+                        num_proc=num_workers,
+                        input_columns=["prompt_input_ids"],
+                    )
+
+    if data_args.save_to_disk is not None and not dataset_was_precomputed:
+        if accelerator.is_main_process:
+            vectorized_datasets.save_to_disk(
+                data_args.save_to_disk,
+                num_proc=min(data_args.preprocessing_num_workers, len(vectorized_datasets["eval"]) - 1),
+            )
+        logger.info(f"Dataset saved at {data_args.save_to_disk}")
+
+    audio_max_length = None
+    if training_args.torch_compile:
+        audio_max_length = max(vectorized_datasets["train"]["target_length"])
+        with accelerator.main_process_first():
+            max_sample = vectorized_datasets["train"].filter(
+                lambda x: x == audio_max_length,
+                num_proc=num_workers,
+                input_columns=["target_length"],
+            )
+        audio_max_length = torch.tensor(max_sample[0]["labels"]).shape[1]
+
+    # for large datasets it is advised to run the preprocessing on a
+    # single machine first with ``args.preprocessing_only`` since there will mostly likely
+    # be a timeout when running the script in distributed mode.
+    # In a second step ``args.preprocessing_only`` can then be set to `False` to load the
+    # cached dataset
+    if data_args.preprocessing_only and data_args.save_to_disk is None:
+        raise ValueError(
+            "`preprocessing_only=True` but `save_to_disk` is not set. The latter should indicates where to save the dataset locally."
+        )
+    elif data_args.preprocessing_only:
+        logger.info(f"Data preprocessing finished. Files save at {data_args.save_to_disk}")
+        return
+
+    # 6. Next, we can prepare the training.
+
+    # Let's use word CLAP similary and WER metrics as our evaluation metrics,
+
+    # Define evaluation metrics during training, *i.e.* CLAP similarity
+    clap = AutoModel.from_pretrained(model_args.clap_model_name_or_path)
+    clap_processor = AutoProcessor.from_pretrained(model_args.clap_model_name_or_path)
+    metric = evaluate.load("wer")
+
+    def clap_similarity(texts, audios, device):
+        clap_inputs = clap_processor(text=texts, audios=audios, padding=True, return_tensors="pt").to(device)
+        clap.to(device)
+        with torch.no_grad():
+            text_features = clap.get_text_features(
+                clap_inputs["input_ids"], attention_mask=clap_inputs.get("attention_mask", None)
+            )
+            audio_features = clap.get_audio_features(clap_inputs["input_features"])
+
+            cosine_sim = torch.nn.functional.cosine_similarity(audio_features, text_features, dim=1, eps=1e-8)
+
+        clap.to("cpu")
+        clap_inputs.to("cpu")
+        return cosine_sim.mean().to("cpu")
+
+    def wer(prompts, audios, device):
+        asr_pipeline = pipeline(model=model_args.asr_model_name_or_path, device=device)
+        transcriptions = asr_pipeline(
+            [{"raw": audio, "sampling_rate": sampling_rate} for audio in audios],
+            batch_size=int(training_args.per_device_eval_batch_size),
+        )
+
+        word_error = 100 * metric.compute(
+            predictions=[t["text"].lower() for t in transcriptions], references=[t.lower() for t in prompts]
+        )
+
+        return word_error, [t["text"] for t in transcriptions]
+
+    eval_methods = {"clap": clap_similarity, "wer": wer}
+
+    def compute_metrics(audios, descriptions, prompts, device="cpu"):
+        input_ids = descriptions
+        texts = description_tokenizer.batch_decode(input_ids, skip_special_tokens=True)
+        prompts = prompt_tokenizer.batch_decode(prompts, skip_special_tokens=True)
+        audios = [a.cpu().numpy() for a in audios]
+        results = {"clap": eval_methods["clap"](texts, audios, device)}
+        word_error, transcriptions = eval_methods["wer"](prompts, audios, device)
+        results["wer"] = word_error
+
+        return results, texts, prompts, audios, transcriptions
+
+    # Define Training Schedule
+    # Store some constants
+    per_device_train_batch_size = int(training_args.per_device_train_batch_size)
+    train_batch_size = per_device_train_batch_size * accelerator.num_processes
+    gradient_accumulation_steps = int(training_args.gradient_accumulation_steps)
+    per_device_eval_batch_size = int(training_args.per_device_eval_batch_size)
+
+    if training_args.max_steps < 0:
+        num_epochs = int(training_args.num_train_epochs)
+        steps_per_epoch = len(vectorized_datasets["train"]) // (train_batch_size * gradient_accumulation_steps)
+        total_train_steps = steps_per_epoch * num_epochs
+    elif training_args.max_steps > 0:
+        logger.info("max_steps is given, it will override any value given in num_train_epochs")
+        total_train_steps = int(training_args.max_steps)
+        # Setting a very large number of epochs so we go as many times as necessary over the iterator.
+        num_epochs = sys.maxsize
+        steps_per_epoch = total_train_steps
+
+    if training_args.evaluation_strategy == "epoch":
+        eval_steps = steps_per_epoch
+    elif training_args.eval_steps is None:
+        logger.info(f"eval_steps is not set, evaluating at the end of each epoch")
+        eval_steps = steps_per_epoch
+    else:
+        eval_steps = training_args.eval_steps
+
+    if training_args.save_strategy == "epoch":
+        save_steps = steps_per_epoch
+    elif training_args.save_strategy == "steps":
+        save_steps = training_args.save_steps
+    else:
+        save_steps = sys.maxsize
+
+    # T5 doesn't support fp16
+    autocast_kwargs = AutocastKwargs(enabled=(mixed_precision != "fp16"))
+
+    # Define optimizer, LR scheduler, collator
+    optimizer = torch.optim.AdamW(
+        params=model.parameters(),
+        lr=training_args.learning_rate,
+        betas=(training_args.adam_beta1, training_args.adam_beta2),
+        eps=training_args.adam_epsilon,
+        weight_decay=training_args.weight_decay,
+    )
+
+    # LR scheduler gets stepped by `num_processes` each time -> account for this in warmup / total steps
+    lr_scheduler = get_scheduler(
+        name=training_args.lr_scheduler_type,
+        optimizer=optimizer,
+        num_warmup_steps=training_args.get_warmup_steps(total_train_steps) * accelerator.num_processes,
+        num_training_steps=total_train_steps * accelerator.num_processes,
+    )
+
+    # Instantiate custom data collator
+    data_collator = DataCollatorParlerTTSWithPadding(
+        prompt_tokenizer=prompt_tokenizer,
+        description_tokenizer=description_tokenizer,
+        pad_to_multiple_of=data_args.pad_to_multiple_of,
+        padding=padding,
+        prompt_max_length=data_args.max_prompt_token_length,
+        description_max_length=data_args.max_description_token_length,
+        audio_max_length=audio_max_length,
+    )
+
+    # Prepare everything with accelerate
+    model, optimizer, lr_scheduler = accelerator.prepare(model, optimizer, lr_scheduler)
+
+    logger.info("***** Running training *****")
+    logger.info(f"  Num examples = {total_train_steps * train_batch_size * gradient_accumulation_steps}")
+    logger.info("  Instantaneous batch size per device =" f" {per_device_train_batch_size}")
+    logger.info("  Gradient accumulation steps =" f" {gradient_accumulation_steps}")
+    logger.info(
+        f"  Total train batch size (w. parallel & distributed) = {train_batch_size * gradient_accumulation_steps}"
+    )
+    logger.info(f"  Total optimization steps = {total_train_steps}")
+
+    # ======================== Training ================================
+    train_time = 0
+    train_start = time.time()
+    steps_trained_progress_bar = tqdm(
+        range(total_train_steps), desc="Train steps ... ", position=0, disable=not accelerator.is_local_main_process
+    )
+    continue_training = True
+    epochs_trained = 0
+    cur_step = 0
+
+    checkpoint = None
+    if training_args.resume_from_checkpoint is not None:
+        checkpoint = training_args.resume_from_checkpoint
+    elif last_checkpoint is not None:
+        checkpoint = last_checkpoint
+
+    if accelerator.is_main_process:
+        if training_args.push_to_hub:
+            # Retrieve of infer repo_name
+            repo_name = training_args.hub_model_id
+            if repo_name is None:
+                repo_name = Path(training_args.output_dir).absolute().name
+            # Create repo and retrieve repo_id
+            repo_id = create_repo(repo_name, exist_ok=True, token=training_args.hub_token).repo_id
+            # Clone repo locally
+            repo = Repository(training_args.output_dir, clone_from=repo_id, token=training_args.hub_token)
+
+            with open(os.path.join(training_args.output_dir, ".gitignore"), "w+") as gitignore:
+                if "wandb" not in gitignore:
+                    gitignore.write("wandb\n")
+        elif training_args.output_dir is not None:
+            os.makedirs(training_args.output_dir, exist_ok=True)
+    accelerator.wait_for_everyone()
+
+    # Now save everything to be able to create a single processor later
+    # make sure all processes wait until data is saved
+    with accelerator.main_process_first():
+        # only the main process saves them
+        if accelerator.is_main_process:
+            # save feature extractor, tokenizer and config
+            if (
+                model_args.prompt_tokenizer_name is None
+                and model_args.description_tokenizer_name
+                or (model_args.prompt_tokenizer_name == model_args.description_tokenizer_name)
+            ):
+                prompt_tokenizer.save_pretrained(training_args.output_dir)
+            else:
+                logger.warning(
+                    "Prompt tokenizer ('{model_args.prompt_tokenizer_name}') and description tokenizer ('{model_args.description_tokenizer_name}') are not the same. Saving only the prompt tokenizer."
+                )
+                prompt_tokenizer.save_pretrained(training_args.output_dir)
+
+            feature_extractor.save_pretrained(training_args.output_dir)
+            config.save_pretrained(training_args.output_dir)
+
+    if checkpoint is not None:
+        accelerator.load_state(checkpoint)
+        # Find num steps and epoch from saved state string pattern
+        pattern = r"checkpoint-(\d+)-epoch-(\d+)"
+        match = re.search(pattern, checkpoint)
+        cur_step = int(match.group(1))
+        epochs_trained = int(match.group(2))
+
+        logger.info("  Continuing training from checkpoint, will skip to saved global_step")
+        logger.info(f"  Continuing training from epoch {epochs_trained}")
+        logger.info(f"  Continuing training from global step {cur_step}")
+
+        steps_trained_progress_bar.update(cur_step)
+
+        for epoch in range(0, epochs_trained):
+            vectorized_datasets["train"] = vectorized_datasets["train"].shuffle(training_args.seed)
+
+        if training_args.max_steps < 0:
+            # we know exactly the number of steps per epoch, so can skip through the required number of batches
+            resume_step = (cur_step - epochs_trained * steps_per_epoch) * gradient_accumulation_steps
+        else:
+            # Currently we don't know how many steps we've taken in the current epoch
+            # So we just shuffle the dataset one extra time and start from a fresh epoch
+            # This is "good enough" for our purposes but not fully correct
+            resume_step = None
+            vectorized_datasets["train"] = vectorized_datasets["train"].shuffle(training_args.seed)
+    else:
+        resume_step = None
+
+    gen_kwargs = {
+        "do_sample": model_args.do_sample,
+        "temperature": model_args.temperature,
+        "max_length": model_args.max_length,
+    }
+
+    # Define gradient update step fn
+    def train_step(
+        batch,
+        accelerator,
+        autocast_kwargs,
+    ):
+        model.train()
+
+        if mixed_precision == "fp16":
+            # fp16 doesn't work with T5-like models
+            with accelerator.autocast(autocast_handler=autocast_kwargs):
+                if training_args.parallel_mode.value != "distributed":
+                    encoder_outputs = model.text_encoder(
+                        input_ids=batch.get("input_ids"), attention_mask=batch.get("attention_mask", None)
+                    )
+                else:
+                    encoder_outputs = model.module.text_encoder(
+                        input_ids=batch.get("input_ids"), attention_mask=batch.get("attention_mask", None)
+                    )
+                batch["encoder_outputs"] = encoder_outputs
+
+        outputs = model(**batch)
+        # CE (data) loss
+        ce_loss = outputs.loss
+
+        metrics = {"loss": ce_loss}
+        return ce_loss, metrics
+
+    # Define eval fn
+    def eval_step(
+        batch,
+        accelerator,
+        autocast_kwargs,
+    ):
+        eval_model = model if not training_args.torch_compile else model._orig_mod
+        eval_model.eval()
+
+        if mixed_precision == "fp16":
+            # fp16 doesn't work with T5-like models
+            with accelerator.autocast(autocast_handler=autocast_kwargs):
+                with torch.no_grad():
+                    if training_args.parallel_mode.value != "distributed" or training_args.torch_compile:
+                        encoder_outputs = eval_model.text_encoder(
+                            input_ids=batch.get("input_ids"), attention_mask=batch.get("attention_mask", None)
+                        )
+                    else:
+                        encoder_outputs = eval_model.module.text_encoder(
+                            input_ids=batch.get("input_ids"), attention_mask=batch.get("attention_mask", None)
+                        )
+                batch["encoder_outputs"] = encoder_outputs
+
+        with torch.no_grad():
+            outputs = eval_model(**batch)
+        # CE (data) loss
+        ce_loss = outputs.loss
+        metrics = {"loss": ce_loss}
+        return metrics
+
+    def generate_step(batch):
+        batch.pop("decoder_attention_mask", None)
+        eval_model = accelerator.unwrap_model(model, keep_fp32_wrapper=mixed_precision != "fp16").eval()
+        if training_args.torch_compile:
+            eval_model = model._orig_mod
+
+        output_audios = eval_model.generate(**batch, **gen_kwargs)
+        output_audios = accelerator.pad_across_processes(output_audios, dim=1, pad_index=0)
+        return output_audios
+
+    for epoch in range(epochs_trained, num_epochs):
+        vectorized_datasets["train"] = vectorized_datasets["train"].shuffle(training_args.seed)
+        sampler = None
+        if training_args.group_by_length:
+            sampler = LengthGroupedSampler(train_batch_size, lengths=vectorized_datasets["train"]["target_length"])
+        train_dataloader = DataLoader(
+            vectorized_datasets["train"],
+            collate_fn=data_collator,
+            batch_size=per_device_train_batch_size,
+            sampler=sampler,
+            num_workers=training_args.dataloader_num_workers,
+            pin_memory=training_args.dataloader_pin_memory,
+        )
+        train_dataloader = accelerator.prepare(train_dataloader)
+        if hasattr(train_dataloader, "dataset") and isinstance(train_dataloader.dataset, IterableDataset):
+            train_dataloader.dataset.set_epoch(epoch)
+
+        if resume_step is not None:
+            # Skip the first N batches in the dataloader when resuming from a checkpoint
+            train_dataloader = accelerator.skip_first_batches(train_dataloader, resume_step)
+            resume_step = None
+
+        for batch in train_dataloader:
+            with accelerator.accumulate(model):
+                loss, train_metric = train_step(batch, accelerator, autocast_kwargs)
+                accelerator.backward(loss)
+                if accelerator.sync_gradients:
+                    accelerator.clip_grad_norm_(model.parameters(), training_args.max_grad_norm)
+                optimizer.step()
+                lr_scheduler.step()
+                optimizer.zero_grad()
+
+            # Check if the accelerator has performed an optimization step behind the scenes
+            if accelerator.sync_gradients:
+                steps_trained_progress_bar.update(1)
+                cur_step += 1
+
+                if cur_step % training_args.logging_steps == 0:
+                    steps_trained_progress_bar.write(
+                        f"Step... ({cur_step} / {total_train_steps} | Loss:"
+                        f" {train_metric['loss']}, Learning Rate:"
+                        f" {lr_scheduler.get_last_lr()[0]})"
+                    )
+                    log_metric(
+                        accelerator,
+                        metrics=train_metric,
+                        learning_rate=lr_scheduler.get_last_lr()[0],
+                        train_time=train_time + time.time() - train_start,
+                        step=cur_step,
+                        epoch=epoch + (cur_step - epoch * steps_per_epoch) / steps_per_epoch,
+                        prefix="train",
+                    )
+
+                # save checkpoint and weights after each save_steps and at the end of training
+                if (cur_step % save_steps == 0) or cur_step == total_train_steps:
+                    intermediate_dir = os.path.join(training_args.output_dir, f"checkpoint-{cur_step}-epoch-{epoch}")
+                    # safe_serialization=False to avoid shared tensors saving issue (TODO(YL): it's a temporary fix)
+                    # https://github.com/huggingface/transformers/issues/27293#issuecomment-1872560074
+                    accelerator.save_state(output_dir=intermediate_dir, safe_serialization=False)
+                    accelerator.wait_for_everyone()
+                    if accelerator.is_main_process:
+                        rotate_checkpoints(training_args.save_total_limit, output_dir=training_args.output_dir)
+
+                        if cur_step == total_train_steps:
+                            # un-wrap student model for save
+                            unwrapped_model = accelerator.unwrap_model(model)
+                            unwrapped_model.save_pretrained(training_args.output_dir)
+
+                        if training_args.push_to_hub:
+                            repo.push_to_hub(
+                                commit_message=f"Saving train state of step {cur_step}",
+                                blocking=False,
+                            )
+
+                if training_args.do_eval and (cur_step % eval_steps == 0 or cur_step == total_train_steps):
+                    train_time += time.time() - train_start
+                    # ======================== Evaluating ==============================
+                    eval_metrics = []
+                    eval_preds = []
+                    eval_descriptions = []
+                    eval_prompts = []
+                    eval_start = time.time()
+
+                    # release training input batch
+                    batch = release_memory(batch)
+
+                    validation_dataloader = DataLoader(
+                        vectorized_datasets["eval"],
+                        collate_fn=data_collator,
+                        batch_size=per_device_eval_batch_size,
+                        drop_last=False,
+                        num_workers=training_args.dataloader_pin_memory,
+                        pin_memory=training_args.dataloader_pin_memory,
+                    )
+                    validation_dataloader = accelerator.prepare(validation_dataloader)
+
+                    for batch in tqdm(
+                        validation_dataloader,
+                        desc="Evaluating - Inference ...",
+                        position=2,
+                        disable=not accelerator.is_local_main_process,
+                    ):
+                        # Model forward
+                        eval_metric = eval_step(batch, accelerator, autocast_kwargs)
+                        eval_metric = accelerator.gather_for_metrics(eval_metric)
+                        eval_metrics.append(eval_metric)
+
+                    if training_args.predict_with_generate:
+                        validation_dataloader = DataLoader(
+                            vectorized_datasets["eval"],
+                            collate_fn=data_collator,
+                            batch_size=per_device_eval_batch_size,
+                            drop_last=False,
+                            num_workers=training_args.dataloader_pin_memory,
+                            pin_memory=training_args.dataloader_pin_memory,
+                        )
+                        validation_dataloader = accelerator.prepare(validation_dataloader)
+                        # generation
+                        for batch in tqdm(
+                            validation_dataloader,
+                            desc="Evaluating - Generation ...",
+                            position=2,
+                            disable=not accelerator.is_local_main_process,
+                        ):
+                            generated_audios = generate_step(batch)
+                            # Gather all predictions and targets
+                            generated_audios, input_ids, prompts = accelerator.pad_across_processes(
+                                (generated_audios, batch["input_ids"], batch["prompt_input_ids"]), dim=1, pad_index=0
+                            )
+                            generated_audios, input_ids, prompts = accelerator.gather_for_metrics(
+                                (generated_audios, input_ids, prompts)
+                            )
+                            eval_preds.extend(generated_audios.to("cpu"))
+                            eval_descriptions.extend(input_ids.to("cpu"))
+                            eval_prompts.extend(prompts.to("cpu"))
+
+                    eval_time = time.time() - eval_start
+                    # normalize eval metrics
+                    eval_metrics = {
+                        key: torch.mean(torch.cat([d[key].unsqueeze(0) for d in eval_metrics]))
+                        for key in eval_metrics[0]
+                    }
+
+                    # compute metrics
+                    metrics_desc = ""
+                    if training_args.predict_with_generate:
+                        metric_values, pred_descriptions, pred_prompts, audios, transcriptions = compute_metrics(
+                            eval_preds, eval_descriptions, eval_prompts, accelerator.device
+                        )
+                        eval_metrics.update(metric_values)
+                        metrics_desc = " ".join([f"Eval {key}: {value} |" for key, value in metric_values.items()])
+                        if "wandb" in training_args.report_to:
+                            log_pred(
+                                accelerator,
+                                pred_descriptions,
+                                pred_prompts,
+                                transcriptions,
+                                audios,
+                                sampling_rate=sampling_rate,
+                                step=cur_step,
+                                prefix="eval",
+                            )
+
+                    # Print metrics and update progress bar
+                    steps_trained_progress_bar.write(
+                        f"Eval results for step ({cur_step} / {total_train_steps} | Eval Loss: {eval_metrics['loss']} |"
+                        f" {metrics_desc})"
+                    )
+
+                    log_metric(
+                        accelerator,
+                        metrics=eval_metrics,
+                        train_time=eval_time,
+                        step=cur_step,
+                        epoch=epoch + (cur_step - epoch * steps_per_epoch) / steps_per_epoch,
+                        prefix="eval",
+                    )
+
+                    # release eval batch and relax metrics
+                    eval_metrics = []
+                    eval_preds = []
+                    eval_descriptions = []
+                    eval_prompts = []
+                    batch = release_memory(batch)
+
+                    # flush the train metrics
+                    train_start = time.time()
+
+                # break condition
+                if cur_step == total_train_steps:
+                    continue_training = False
+                    break
+
+        if not continue_training:
+            break
+
+    accelerator.end_training()
+
+
+if __name__ == "__main__":
+    set_start_method("spawn")
+    main()

wandb/debug.log

@@ -0,0 +1,35 @@
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_setup.py:_flush():76] Current SDK version is 0.17.0
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_setup.py:_flush():76] Configure stats pid to 1257680
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_setup.py:_flush():76] Loading settings from /home/sanchit/.config/wandb/settings
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_setup.py:_flush():76] Loading settings from /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/settings
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_setup.py:_flush():76] Loading settings from environment variables: {}
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_setup.py:_flush():76] Applying setup settings: {'_disable_service': False}
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_setup.py:_flush():76] Inferring run settings from compute environment: {'program_relpath': 'run_parler_tts_training.py', 'program_abspath': '/raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/run_parler_tts_training.py', 'program': '/raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/run_parler_tts_training.py'}
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_setup.py:_flush():76] Applying login settings: {}
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_init.py:_log_setup():520] Logging user logs to /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_205249-qaoje1x9/logs/debug.log
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_init.py:_log_setup():521] Logging internal logs to /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_205249-qaoje1x9/logs/debug-internal.log
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_init.py:init():560] calling init triggers
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_init.py:init():567] wandb.init called with sweep_config: {}
+config: {}
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_init.py:init():610] starting backend
+2024-05-13 20:52:49,015 INFO    MainThread:1257680 [wandb_init.py:init():614] setting up manager
+2024-05-13 20:52:49,019 INFO    MainThread:1257680 [backend.py:_multiprocessing_setup():105] multiprocessing start_methods=fork,spawn,forkserver, using: spawn
+2024-05-13 20:52:49,019 INFO    MainThread:1257680 [wandb_init.py:init():622] backend started and connected
+2024-05-13 20:52:49,021 INFO    MainThread:1257680 [wandb_init.py:init():711] updated telemetry
+2024-05-13 20:52:49,024 INFO    MainThread:1257680 [wandb_init.py:init():744] communicating run to backend with 90.0 second timeout
+2024-05-13 20:52:49,413 INFO    MainThread:1257680 [wandb_run.py:_on_init():2396] communicating current version
+2024-05-13 20:52:49,474 INFO    MainThread:1257680 [wandb_run.py:_on_init():2405] got version response 
+2024-05-13 20:52:49,475 INFO    MainThread:1257680 [wandb_init.py:init():795] starting run threads in backend
+2024-05-13 20:52:52,103 INFO    MainThread:1257680 [wandb_run.py:_console_start():2374] atexit reg
+2024-05-13 20:52:52,103 INFO    MainThread:1257680 [wandb_run.py:_redirect():2229] redirect: wrap_raw
+2024-05-13 20:52:52,103 INFO    MainThread:1257680 [wandb_run.py:_redirect():2294] Wrapping output streams.
+2024-05-13 20:52:52,103 INFO    MainThread:1257680 [wandb_run.py:_redirect():2319] Redirects installed.
+2024-05-13 20:52:52,104 INFO    MainThread:1257680 [wandb_init.py:init():838] run started, returning control to user process
+2024-05-13 20:52:52,104 INFO    MainThread:1257680 [wandb_run.py:_config_callback():1376] config_cb None None {'learning_rate': 8e-05, 'model_name_or_path': 'parler-tts/parler_tts_mini_v0.1', 'num_train_epochs': 8, 'gradient_accumulation_steps': 8, 'per_device_train_batch_size': 16, 'global_batch_size': 16, 'mixed_precision': 'bf16', 'lr_scheduler_type': 'SchedulerType.COSINE', 'warmup_steps': 250, 'freeze_text_encoder': True, 'max_duration_in_seconds': 30.0, 'weight_decay': 0.01, 'adam_beta1': 0.9, 'adam_beta2': 0.99, 'temperature': 1.0}
+2024-05-13 23:01:32,489 INFO    MainThread:1257680 [wandb_run.py:_finish():2103] finishing run sanchit-gandhi/parler-speech/qaoje1x9
+2024-05-13 23:01:32,489 INFO    MainThread:1257680 [wandb_run.py:_atexit_cleanup():2343] got exitcode: 0
+2024-05-13 23:01:32,489 INFO    MainThread:1257680 [wandb_run.py:_restore():2326] restore
+2024-05-13 23:01:32,489 INFO    MainThread:1257680 [wandb_run.py:_restore():2332] restore done
+2024-05-13 23:01:46,253 INFO    MainThread:1257680 [wandb_run.py:_footer_history_summary_info():3994] rendering history
+2024-05-13 23:01:46,254 INFO    MainThread:1257680 [wandb_run.py:_footer_history_summary_info():4026] rendering summary
+2024-05-13 23:01:46,256 INFO    MainThread:1257680 [wandb_run.py:_footer_sync_info():3953] logging synced files

wandb/run-20240513_204652-m0g0ap7d/files/conda-environment.yaml

@@ -0,0 +1,248 @@
+name: venv
+channels:
+  - defaults
+dependencies:
+  - _libgcc_mutex=0.1=main
+  - _openmp_mutex=5.1=1_gnu
+  - bzip2=1.0.8=h5eee18b_6
+  - ca-certificates=2024.3.11=h06a4308_0
+  - ld_impl_linux-64=2.38=h1181459_1
+  - libffi=3.4.4=h6a678d5_1
+  - libgcc-ng=11.2.0=h1234567_1
+  - libgomp=11.2.0=h1234567_1
+  - libstdcxx-ng=11.2.0=h1234567_1
+  - libuuid=1.41.5=h5eee18b_0
+  - ncurses=6.4=h6a678d5_0
+  - openssl=3.0.13=h7f8727e_1
+  - pip=24.0=py311h06a4308_0
+  - python=3.11.9=h955ad1f_0
+  - readline=8.2=h5eee18b_0
+  - setuptools=69.5.1=py311h06a4308_0
+  - sqlite=3.45.3=h5eee18b_0
+  - tk=8.6.14=h39e8969_0
+  - wheel=0.43.0=py311h06a4308_0
+  - xz=5.4.6=h5eee18b_1
+  - zlib=1.2.13=h5eee18b_1
+  - pip:
+      - absl-py==2.1.0
+      - accelerate==0.30.0
+      - aiohttp==3.9.5
+      - aiosignal==1.3.1
+      - aniso8601==9.0.1
+      - annotated-types==0.6.0
+      - anyio==4.3.0
+      - argbind==0.3.7
+      - argon2-cffi==23.1.0
+      - argon2-cffi-bindings==21.2.0
+      - arrow==1.3.0
+      - asttokens==2.4.1
+      - async-lru==2.0.4
+      - attrs==23.2.0
+      - audioread==3.0.1
+      - babel==2.15.0
+      - beautifulsoup4==4.12.3
+      - bidict==0.23.1
+      - bitsandbytes==0.43.1
+      - bleach==6.1.0
+      - certifi==2024.2.2
+      - cffi==1.16.0
+      - charset-normalizer==3.3.2
+      - click==8.1.7
+      - coloredlogs==14.0
+      - comm==0.2.2
+      - contourpy==1.2.1
+      - cycler==0.12.1
+      - datasets==2.19.1
+      - debugpy==1.8.1
+      - decorator==5.1.1
+      - defusedxml==0.7.1
+      - descript-audio-codec==1.0.0
+      - descript-audiotools==0.7.2
+      - dill==0.3.8
+      - dnspython==2.3.0
+      - docker-pycreds==0.4.0
+      - docstring-parser==0.16
+      - editdistance==0.8.1
+      - einops==0.8.0
+      - et-xmlfile==1.1.0
+      - evaluate==0.4.2
+      - eventlet==0.36.1
+      - executing==2.0.1
+      - fastjsonschema==2.19.1
+      - ffmpy==0.3.2
+      - filelock==3.14.0
+      - fire==0.6.0
+      - flask==2.2.5
+      - flask-cors==4.0.1
+      - flask-restful==0.3.10
+      - flask-socketio==5.3.6
+      - flask-talisman==1.1.0
+      - flatten-dict==0.4.2
+      - fonttools==4.51.0
+      - fqdn==1.5.1
+      - frozenlist==1.4.1
+      - fsspec==2024.3.1
+      - future==1.0.0
+      - g2p==2.0.0
+      - gitdb==4.0.11
+      - gitpython==3.1.43
+      - greenlet==3.0.3
+      - grpcio==1.63.0
+      - h11==0.14.0
+      - httpcore==1.0.5
+      - httpx==0.27.0
+      - huggingface-hub==0.23.0
+      - humanfriendly==10.0
+      - idna==3.7
+      - importlib-resources==6.4.0
+      - ipdb==0.13.13
+      - ipykernel==6.29.4
+      - ipython==8.24.0
+      - isoduration==20.11.0
+      - itsdangerous==2.2.0
+      - jedi==0.19.1
+      - jinja2==3.1.4
+      - jiwer==3.0.4
+      - joblib==1.4.2
+      - json5==0.9.25
+      - jsonpointer==2.4
+      - jsonschema==4.22.0
+      - jsonschema-specifications==2023.12.1
+      - julius==0.2.7
+      - jupyter-client==8.6.1
+      - jupyter-core==5.7.2
+      - jupyter-events==0.10.0
+      - jupyter-lsp==2.2.5
+      - jupyter-server==2.14.0
+      - jupyter-server-terminals==0.5.3
+      - jupyterlab==4.2.0
+      - jupyterlab-pygments==0.3.0
+      - jupyterlab-server==2.27.1
+      - kiwisolver==1.4.5
+      - lazy-loader==0.4
+      - librosa==0.10.2
+      - llvmlite==0.42.0
+      - markdown==3.6
+      - markdown-it-py==3.0.0
+      - markdown2==2.4.13
+      - markupsafe==2.1.5
+      - matplotlib==3.8.4
+      - matplotlib-inline==0.1.7
+      - mdurl==0.1.2
+      - mistune==3.0.2
+      - mpmath==1.3.0
+      - msgpack==1.0.8
+      - multidict==6.0.5
+      - multiprocess==0.70.16
+      - munkres==1.1.4
+      - nbclient==0.10.0
+      - nbconvert==7.16.4
+      - nbformat==5.10.4
+      - nest-asyncio==1.6.0
+      - networkx==3.3
+      - notebook-shim==0.2.4
+      - numba==0.59.1
+      - numpy==1.26.4
+      - nvidia-cublas-cu12==12.1.3.1
+      - nvidia-cuda-cupti-cu12==12.1.105
+      - nvidia-cuda-nvrtc-cu12==12.1.105
+      - nvidia-cuda-runtime-cu12==12.1.105
+      - nvidia-cudnn-cu12==8.9.2.26
+      - nvidia-cufft-cu12==11.0.2.54
+      - nvidia-curand-cu12==10.3.2.106
+      - nvidia-cusolver-cu12==11.4.5.107
+      - nvidia-cusparse-cu12==12.1.0.106
+      - nvidia-nccl-cu12==2.20.5
+      - nvidia-nvjitlink-cu12==12.4.127
+      - nvidia-nvtx-cu12==12.1.105
+      - openpyxl==3.1.2
+      - overrides==7.7.0
+      - packaging==24.0
+      - pandas==2.2.2
+      - pandocfilters==1.5.1
+      - panphon==0.20.0
+      - parler-tts==0.1
+      - parso==0.8.4
+      - pexpect==4.9.0
+      - pillow==10.3.0
+      - platformdirs==4.2.1
+      - pooch==1.8.1
+      - prometheus-client==0.20.0
+      - prompt-toolkit==3.0.43
+      - protobuf==3.19.6
+      - psutil==5.9.8
+      - ptyprocess==0.7.0
+      - pure-eval==0.2.2
+      - pyarrow==16.0.0
+      - pyarrow-hotfix==0.6
+      - pycparser==2.22
+      - pydantic==2.7.1
+      - pydantic-core==2.18.2
+      - pygments==2.18.0
+      - pyloudnorm==0.1.1
+      - pyparsing==3.1.2
+      - pystoi==0.4.1
+      - python-dateutil==2.9.0.post0
+      - python-engineio==4.9.0
+      - python-json-logger==2.0.7
+      - python-socketio==5.11.2
+      - pytz==2024.1
+      - pyyaml==6.0.1
+      - pyzmq==26.0.3
+      - randomname==0.2.1
+      - rapidfuzz==3.9.0
+      - referencing==0.35.1
+      - regex==2024.4.28
+      - requests==2.31.0
+      - rfc3339-validator==0.1.4
+      - rfc3986-validator==0.1.1
+      - rich==13.7.1
+      - rpds-py==0.18.1
+      - safetensors==0.4.3
+      - scikit-learn==1.4.2
+      - scipy==1.13.0
+      - send2trash==1.8.3
+      - sentencepiece==0.2.0
+      - sentry-sdk==2.1.1
+      - setproctitle==1.3.3
+      - simple-websocket==1.0.0
+      - six==1.16.0
+      - smmap==5.0.1
+      - sniffio==1.3.1
+      - soundfile==0.12.1
+      - soupsieve==2.5
+      - soxr==0.3.7
+      - stack-data==0.6.3
+      - sympy==1.12
+      - tensorboard==2.16.2
+      - tensorboard-data-server==0.7.2
+      - termcolor==2.4.0
+      - terminado==0.18.1
+      - text-unidecode==1.3
+      - threadpoolctl==3.5.0
+      - tinycss2==1.3.0
+      - tokenizers==0.19.1
+      - torch==2.3.0
+      - torch-stoi==0.2.1
+      - torchaudio==2.3.0
+      - tornado==6.4
+      - tqdm==4.66.4
+      - traitlets==5.14.3
+      - transformers==4.41.0.dev0
+      - triton==2.3.0
+      - types-python-dateutil==2.9.0.20240316
+      - typing-extensions==4.11.0
+      - tzdata==2024.1
+      - unicodecsv==0.14.1
+      - uri-template==1.3.0
+      - urllib3==2.2.1
+      - wandb==0.17.0
+      - wcwidth==0.2.13
+      - webcolors==1.13
+      - webencodings==0.5.1
+      - websocket-client==1.8.0
+      - werkzeug==3.0.3
+      - wsproto==1.2.0
+      - xxhash==3.4.1
+      - yarl==1.9.4
+prefix: /home/sanchit/miniconda3/envs/venv

wandb/run-20240513_204652-m0g0ap7d/files/config.yaml

@@ -0,0 +1,86 @@
+wandb_version: 1
+
+_wandb:
+  desc: null
+  value:
+    python_version: 3.11.9
+    cli_version: 0.17.0
+    framework: huggingface
+    huggingface_version: 4.41.0.dev0
+    is_jupyter_run: false
+    is_kaggle_kernel: false
+    start_time: 1715626012
+    t:
+      1:
+      - 1
+      - 5
+      - 11
+      - 49
+      - 51
+      - 53
+      - 55
+      - 71
+      - 100
+      2:
+      - 1
+      - 5
+      - 11
+      - 49
+      - 51
+      - 53
+      - 55
+      - 71
+      - 100
+      3:
+      - 23
+      4: 3.11.9
+      5: 0.17.0
+      6: 4.41.0.dev0
+      8:
+      - 5
+      13: linux-x86_64
+learning_rate:
+  desc: null
+  value: 8.0e-05
+model_name_or_path:
+  desc: null
+  value: parler-tts/parler_tts_mini_v0.1
+num_train_epochs:
+  desc: null
+  value: 8
+gradient_accumulation_steps:
+  desc: null
+  value: 8
+per_device_train_batch_size:
+  desc: null
+  value: 16
+global_batch_size:
+  desc: null
+  value: 16
+mixed_precision:
+  desc: null
+  value: bf16
+lr_scheduler_type:
+  desc: null
+  value: SchedulerType.COSINE
+warmup_steps:
+  desc: null
+  value: 250
+freeze_text_encoder:
+  desc: null
+  value: true
+max_duration_in_seconds:
+  desc: null
+  value: 30.0
+weight_decay:
+  desc: null
+  value: 0.01
+adam_beta1:
+  desc: null
+  value: 0.9
+adam_beta2:
+  desc: null
+  value: 0.99
+temperature:
+  desc: null
+  value: 1.0

wandb/run-20240513_204652-m0g0ap7d/files/output.log

@@ -0,0 +1,180 @@
+05/13/2024 20:46:55 - WARNING - __main__ - Process rank: 0, device: cuda:0, n_gpu: 1, distributed training: False, 16-bits training: False
+05/13/2024 20:46:55 - INFO - __main__ - Training/evaluation parameters ParlerTTSTrainingArguments(
+_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},
+adafactor=False,
+adam_beta1=0.9,
+adam_beta2=0.99,
+adam_epsilon=1e-08,
+audio_encoder_per_device_batch_size=4,
+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=4,
+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=1800,
+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_steps=None,
+eval_strategy=IntervalStrategy.EPOCH,
+evaluation_strategy=epoch,
+fp16=False,
+fp16_backend=auto,
+fp16_full_eval=False,
+fp16_opt_level=O1,
+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=8,
+gradient_checkpointing=True,
+gradient_checkpointing_kwargs=None,
+greater_is_better=None,
+group_by_length=True,
+half_precision_backend=auto,
+hub_always_push=False,
+hub_model_id=None,
+hub_private_repo=False,
+hub_strategy=HubStrategy.EVERY_SAVE,
+hub_token=<HUB_TOKEN>,
+ignore_data_skip=False,
+include_inputs_for_metrics=True,
+include_num_input_tokens_seen=False,
+include_tokens_per_second=False,
+jit_mode_eval=False,
+label_names=None,
+label_smoothing_factor=0.0,
+learning_rate=8e-05,
+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=../output_dir_training_constant_concat/runs/May13_20-46-51_hf-dgx-01,
+logging_first_step=False,
+logging_nan_inf_filter=True,
+logging_steps=5,
+logging_strategy=IntervalStrategy.STEPS,
+lr_scheduler_kwargs={},
+lr_scheduler_type=SchedulerType.COSINE,
+max_grad_norm=1.0,
+max_steps=-1,
+metric_for_best_model=None,
+mp_parameters=,
+neftune_noise_alpha=None,
+no_cuda=False,
+num_train_epochs=8,
+optim=OptimizerNames.ADAMW_TORCH,
+optim_args=None,
+optim_target_modules=None,
+output_dir=../output_dir_training_constant_concat/,
+overwrite_output_dir=True,
+past_index=-1,
+per_device_eval_batch_size=16,
+per_device_train_batch_size=16,
+predict_with_generate=True,
+prediction_loss_only=False,
+push_to_hub=False,
+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=['wandb'],
+restore_callback_states_from_checkpoint=False,
+resume_from_checkpoint=None,
+run_name=../output_dir_training_constant_concat/,
+save_on_each_node=False,
+save_only_model=False,
+save_safetensors=True,
+save_steps=500,
+save_strategy=IntervalStrategy.EPOCH,
+save_total_limit=5,
+seed=456,
+skip_memory_metrics=True,
+sortish_sampler=False,
+split_batches=None,
+tf32=None,
+torch_compile=False,
+torch_compile_backend=None,
+torch_compile_mode=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=250,
+weight_decay=0.01,
+)
+05/13/2024 20:46:57 - WARNING - __main__ - Disabling fast tokenizer warning: https://github.com/huggingface/transformers/blob/main/src/transformers/tokenization_utils_base.py#L3231-L3235
+loading configuration file preprocessor_config.json from cache at /raid/.cache/huggingface/models--parler-tts--dac_44khZ_8kbps/snapshots/db52bea859d9411e0beb44a3ea923a8731ee4197/preprocessor_config.json
+Feature extractor EncodecFeatureExtractor {
+  "chunk_length_s": null,
+  "feature_extractor_type": "EncodecFeatureExtractor",
+  "feature_size": 1,
+  "overlap": null,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "return_attention_mask": true,
+  "sampling_rate": 44100
+}
+loading file spiece.model from cache at /raid/.cache/huggingface/models--parler-tts--parler_tts_mini_v0.1/snapshots/e02fd18e77d38b49a85c7a9a85189a64b8472544/spiece.model
+loading file tokenizer.json from cache at /raid/.cache/huggingface/models--parler-tts--parler_tts_mini_v0.1/snapshots/e02fd18e77d38b49a85c7a9a85189a64b8472544/tokenizer.json
+loading file added_tokens.json from cache at None
+loading file special_tokens_map.json from cache at /raid/.cache/huggingface/models--parler-tts--parler_tts_mini_v0.1/snapshots/e02fd18e77d38b49a85c7a9a85189a64b8472544/special_tokens_map.json
+loading file tokenizer_config.json from cache at /raid/.cache/huggingface/models--parler-tts--parler_tts_mini_v0.1/snapshots/e02fd18e77d38b49a85c7a9a85189a64b8472544/tokenizer_config.json
+You set `add_prefix_space`. The tokenizer needs to be converted from the slow tokenizers
+loading file spiece.model from cache at /raid/.cache/huggingface/models--parler-tts--parler_tts_mini_v0.1/snapshots/e02fd18e77d38b49a85c7a9a85189a64b8472544/spiece.model
+loading file tokenizer.json from cache at /raid/.cache/huggingface/models--parler-tts--parler_tts_mini_v0.1/snapshots/e02fd18e77d38b49a85c7a9a85189a64b8472544/tokenizer.json
+loading file added_tokens.json from cache at None
+loading file special_tokens_map.json from cache at /raid/.cache/huggingface/models--parler-tts--parler_tts_mini_v0.1/snapshots/e02fd18e77d38b49a85c7a9a85189a64b8472544/special_tokens_map.json
+loading file tokenizer_config.json from cache at /raid/.cache/huggingface/models--parler-tts--parler_tts_mini_v0.1/snapshots/e02fd18e77d38b49a85c7a9a85189a64b8472544/tokenizer_config.json
+Combining datasets...:   0%|                                                                      | 0/4 [00:00<?, ?it/s]
+Combining datasets...:   0%|                                                                      | 0/4 [03:35<?, ?it/s]
+Traceback (most recent call last):
+  File "/raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/run_parler_tts_training.py", line 1763, in <module>
+    main()
+  File "/raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/run_parler_tts_training.py", line 950, in main
+    raw_datasets["train"] = load_multiple_datasets(
+                            ^^^^^^^^^^^^^^^^^^^^^^^
+  File "/raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/run_parler_tts_training.py", line 693, in load_multiple_datasets
+    metadata_dataset = load_dataset(
+                       ^^^^^^^^^^^^^
+  File "/home/sanchit/miniconda3/envs/venv/lib/python3.11/site-packages/datasets/load.py", line 2587, in load_dataset
+    builder_instance = load_dataset_builder(
+                       ^^^^^^^^^^^^^^^^^^^^^
+  File "/home/sanchit/miniconda3/envs/venv/lib/python3.11/site-packages/datasets/load.py", line 2296, in load_dataset_builder
+    builder_instance: DatasetBuilder = builder_cls(
+                                       ^^^^^^^^^^^^
+  File "/home/sanchit/miniconda3/envs/venv/lib/python3.11/site-packages/datasets/builder.py", line 374, in __init__
+    self.config, self.config_id = self._create_builder_config(
+                                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+  File "/home/sanchit/miniconda3/envs/venv/lib/python3.11/site-packages/datasets/builder.py", line 599, in _create_builder_config
+    raise ValueError(
+ValueError: BuilderConfig 'read' not found. Available: ['default']

wandb/run-20240513_204652-m0g0ap7d/files/requirements.txt

@@ -0,0 +1,225 @@
+Babel==2.15.0
+Flask-Cors==4.0.1
+Flask-RESTful==0.3.10
+Flask-SocketIO==5.3.6
+Flask==2.2.5
+GitPython==3.1.43
+Jinja2==3.1.4
+Markdown==3.6
+MarkupSafe==2.1.5
+PyYAML==6.0.1
+Pygments==2.18.0
+Send2Trash==1.8.3
+Werkzeug==3.0.3
+absl-py==2.1.0
+accelerate==0.30.0
+aiohttp==3.9.5
+aiosignal==1.3.1
+aniso8601==9.0.1
+annotated-types==0.6.0
+anyio==4.3.0
+argbind==0.3.7
+argon2-cffi-bindings==21.2.0
+argon2-cffi==23.1.0
+arrow==1.3.0
+asttokens==2.4.1
+async-lru==2.0.4
+attrs==23.2.0
+audioread==3.0.1
+beautifulsoup4==4.12.3
+bidict==0.23.1
+bitsandbytes==0.43.1
+bleach==6.1.0
+certifi==2024.2.2
+cffi==1.16.0
+charset-normalizer==3.3.2
+click==8.1.7
+coloredlogs==14.0
+comm==0.2.2
+contourpy==1.2.1
+cycler==0.12.1
+datasets==2.19.1
+debugpy==1.8.1
+decorator==5.1.1
+defusedxml==0.7.1
+descript-audio-codec==1.0.0
+descript-audiotools==0.7.2
+dill==0.3.8
+dnspython==2.3.0
+docker-pycreds==0.4.0
+docstring_parser==0.16
+editdistance==0.8.1
+einops==0.8.0
+et-xmlfile==1.1.0
+evaluate==0.4.2
+eventlet==0.36.1
+executing==2.0.1
+fastjsonschema==2.19.1
+ffmpy==0.3.2
+filelock==3.14.0
+fire==0.6.0
+flask-talisman==1.1.0
+flatten-dict==0.4.2
+fonttools==4.51.0
+fqdn==1.5.1
+frozenlist==1.4.1
+fsspec==2024.3.1
+future==1.0.0
+g2p==2.0.0
+gitdb==4.0.11
+greenlet==3.0.3
+grpcio==1.63.0
+h11==0.14.0
+httpcore==1.0.5
+httpx==0.27.0
+huggingface-hub==0.23.0
+humanfriendly==10.0
+idna==3.7
+importlib_resources==6.4.0
+ipdb==0.13.13
+ipykernel==6.29.4
+ipython==8.24.0
+isoduration==20.11.0
+itsdangerous==2.2.0
+jedi==0.19.1
+jiwer==3.0.4
+joblib==1.4.2
+json5==0.9.25
+jsonpointer==2.4
+jsonschema-specifications==2023.12.1
+jsonschema==4.22.0
+julius==0.2.7
+jupyter-events==0.10.0
+jupyter-lsp==2.2.5
+jupyter_client==8.6.1
+jupyter_core==5.7.2
+jupyter_server==2.14.0
+jupyter_server_terminals==0.5.3
+jupyterlab==4.2.0
+jupyterlab_pygments==0.3.0
+jupyterlab_server==2.27.1
+kiwisolver==1.4.5
+lazy_loader==0.4
+librosa==0.10.2
+llvmlite==0.42.0
+markdown-it-py==3.0.0
+markdown2==2.4.13
+matplotlib-inline==0.1.7
+matplotlib==3.8.4
+mdurl==0.1.2
+mistune==3.0.2
+mpmath==1.3.0
+msgpack==1.0.8
+multidict==6.0.5
+multiprocess==0.70.16
+munkres==1.1.4
+nbclient==0.10.0
+nbconvert==7.16.4
+nbformat==5.10.4
+nest-asyncio==1.6.0
+networkx==3.3
+notebook_shim==0.2.4
+numba==0.59.1
+numpy==1.26.4
+nvidia-cublas-cu12==12.1.3.1
+nvidia-cuda-cupti-cu12==12.1.105
+nvidia-cuda-nvrtc-cu12==12.1.105
+nvidia-cuda-runtime-cu12==12.1.105
+nvidia-cudnn-cu12==8.9.2.26
+nvidia-cufft-cu12==11.0.2.54
+nvidia-curand-cu12==10.3.2.106
+nvidia-cusolver-cu12==11.4.5.107
+nvidia-cusparse-cu12==12.1.0.106
+nvidia-nccl-cu12==2.20.5
+nvidia-nvjitlink-cu12==12.4.127
+nvidia-nvtx-cu12==12.1.105
+openpyxl==3.1.2
+overrides==7.7.0
+packaging==24.0
+pandas==2.2.2
+pandocfilters==1.5.1
+panphon==0.20.0
+parler_tts==0.1
+parso==0.8.4
+pexpect==4.9.0
+pillow==10.3.0
+pip==24.0
+platformdirs==4.2.1
+pooch==1.8.1
+prometheus_client==0.20.0
+prompt-toolkit==3.0.43
+protobuf==3.19.6
+psutil==5.9.8
+ptyprocess==0.7.0
+pure-eval==0.2.2
+pyarrow-hotfix==0.6
+pyarrow==16.0.0
+pycparser==2.22
+pydantic==2.7.1
+pydantic_core==2.18.2
+pyloudnorm==0.1.1
+pyparsing==3.1.2
+pystoi==0.4.1
+python-dateutil==2.9.0.post0
+python-engineio==4.9.0
+python-json-logger==2.0.7
+python-socketio==5.11.2
+pytz==2024.1
+pyzmq==26.0.3
+randomname==0.2.1
+rapidfuzz==3.9.0
+referencing==0.35.1
+regex==2024.4.28
+requests==2.31.0
+rfc3339-validator==0.1.4
+rfc3986-validator==0.1.1
+rich==13.7.1
+rpds-py==0.18.1
+safetensors==0.4.3
+scikit-learn==1.4.2
+scipy==1.13.0
+sentencepiece==0.2.0
+sentry-sdk==2.1.1
+setproctitle==1.3.3
+setuptools==69.5.1
+simple-websocket==1.0.0
+six==1.16.0
+smmap==5.0.1
+sniffio==1.3.1
+soundfile==0.12.1
+soupsieve==2.5
+soxr==0.3.7
+stack-data==0.6.3
+sympy==1.12
+tensorboard-data-server==0.7.2
+tensorboard==2.16.2
+termcolor==2.4.0
+terminado==0.18.1
+text-unidecode==1.3
+threadpoolctl==3.5.0
+tinycss2==1.3.0
+tokenizers==0.19.1
+torch-stoi==0.2.1
+torch==2.3.0
+torchaudio==2.3.0
+tornado==6.4
+tqdm==4.66.4
+traitlets==5.14.3
+transformers==4.41.0.dev0
+transformers==4.41.0.dev0
+triton==2.3.0
+types-python-dateutil==2.9.0.20240316
+typing_extensions==4.11.0
+tzdata==2024.1
+unicodecsv==0.14.1
+uri-template==1.3.0
+urllib3==2.2.1
+wandb==0.17.0
+wcwidth==0.2.13
+webcolors==1.13
+webencodings==0.5.1
+websocket-client==1.8.0
+wheel==0.43.0
+wsproto==1.2.0
+xxhash==3.4.1
+yarl==1.9.4

wandb/run-20240513_204652-m0g0ap7d/files/wandb-metadata.json

@@ -0,0 +1,706 @@
+{
+    "os": "Linux-5.4.0-166-generic-x86_64-with-glibc2.31",
+    "python": "3.11.9",
+    "heartbeatAt": "2024-05-13T18:46:53.365083",
+    "startedAt": "2024-05-13T18:46:52.816759",
+    "docker": null,
+    "cuda": null,
+    "args": [
+        "finetuning_concatenated_config.json"
+    ],
+    "state": "running",
+    "program": "/raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/run_parler_tts_training.py",
+    "codePathLocal": "run_parler_tts_training.py",
+    "codePath": "run_parler_tts_training.py",
+    "git": {
+        "remote": "https://huggingface.co/sanchit-gandhi/parler-tts-mini-v0.1-expresso-concatenated-combined",
+        "commit": "50ba4323d7b8bb052629aa1b88283b9df081a821"
+    },
+    "email": "[email protected]",
+    "root": "/raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined",
+    "host": "hf-dgx-01",
+    "username": "sanchit",
+    "executable": "/home/sanchit/miniconda3/envs/venv/bin/python",
+    "cpu_count": 64,
+    "cpu_count_logical": 128,
+    "cpu_freq": {
+        "current": 2257.736234375,
+        "min": 1500.0,
+        "max": 2250.0
+    },
+    "cpu_freq_per_core": [
+        {
+            "current": 1795.281,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1794.292,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1795.78,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1792.55,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1742.094,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3026.54,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1786.214,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1742.547,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1728.916,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1734.023,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3195.219,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1733.722,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3341.911,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3325.608,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3233.258,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1669.847,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1791.948,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1796.794,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1791.49,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1793.945,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3342.943,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1669.791,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1669.593,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1694.312,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1873.727,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1724.813,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2354.471,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1718.662,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1670.588,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1665.577,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1616.671,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2080.81,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1670.666,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1652.559,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3323.654,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1671.311,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1726.286,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1670.365,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3320.57,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1669.941,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1791.021,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1796.246,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1793.946,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1794.848,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3339.327,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3344.315,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3338.901,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1668.541,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1794.526,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1792.886,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1796.844,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1793.81,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1724.861,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2294.458,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1720.835,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1720.155,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1668.96,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1976.5,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2241.578,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1671.964,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3319.623,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1670.777,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1670.389,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1669.629,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1794.19,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1794.138,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1796.317,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1792.821,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1794.716,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1793.624,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1796.346,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1793.897,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1735.424,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1738.64,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1979.998,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1737.286,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3313.748,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3337.223,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1671.416,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1670.005,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1794.276,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1738.22,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1742.737,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1770.535,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3320.252,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1671.037,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1669.549,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1670.948,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2843.391,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2348.589,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3287.915,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2340.192,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2426.358,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2415.833,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2419.416,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2277.433,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2365.562,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2400.6,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2075.143,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2382.295,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3066.339,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2466.631,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3100.81,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2421.93,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3233.829,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2234.583,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2452.089,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2975.985,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3301.512,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3336.905,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2984.87,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2384.306,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2965.197,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1929.067,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1986.731,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1999.412,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2477.541,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3111.851,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2009.907,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 1993.784,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2144.459,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3337.426,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3320.114,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2169.719,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 3308.644,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2111.633,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2123.71,
+            "min": 1500.0,
+            "max": 2250.0
+        },
+        {
+            "current": 2153.49,
+            "min": 1500.0,
+            "max": 2250.0
+        }
+    ],
+    "disk": {
+        "/": {
+            "total": 1757.8785285949707,
+            "used": 1663.5005989074707
+        }
+    },
+    "gpu": "NVIDIA A100-SXM4-80GB",
+    "gpu_count": 5,
+    "gpu_devices": [
+        {
+            "name": "NVIDIA A100-SXM4-80GB",
+            "memory_total": 85899345920
+        },
+        {
+            "name": "NVIDIA A100-SXM4-80GB",
+            "memory_total": 85899345920
+        },
+        {
+            "name": "NVIDIA A100-SXM4-80GB",
+            "memory_total": 85899345920
+        },
+        {
+            "name": "NVIDIA DGX Display",
+            "memory_total": 4294967296
+        },
+        {
+            "name": "NVIDIA A100-SXM4-80GB",
+            "memory_total": 85899345920
+        }
+    ],
+    "memory": {
+        "total": 503.5396919250488
+    }
+}

wandb/run-20240513_204652-m0g0ap7d/files/wandb-summary.json

@@ -0,0 +1 @@
+{"_wandb": {"runtime": 219}}

wandb/run-20240513_204652-m0g0ap7d/logs/debug-internal.log

@@ -0,0 +1,455 @@
+2024-05-13 20:46:52,823 INFO    StreamThr :1244775 [internal.py:wandb_internal():85] W&B internal server running at pid: 1244775, started at: 2024-05-13 20:46:52.823043
+2024-05-13 20:46:52,825 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status
+2024-05-13 20:46:52,826 INFO    WriterThread:1244775 [datastore.py:open_for_write():87] open: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/run-m0g0ap7d.wandb
+2024-05-13 20:46:52,828 DEBUG   SenderThread:1244775 [sender.py:send():378] send: header
+2024-05-13 20:46:52,829 DEBUG   SenderThread:1244775 [sender.py:send():378] send: run
+2024-05-13 20:46:53,223 INFO    SenderThread:1244775 [dir_watcher.py:__init__():211] watching files in: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files
+2024-05-13 20:46:53,223 INFO    SenderThread:1244775 [sender.py:_start_run_threads():1123] run started: m0g0ap7d with start time 1715626012.822173
+2024-05-13 20:46:53,229 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: check_version
+2024-05-13 20:46:53,229 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: check_version
+2024-05-13 20:46:53,293 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: run_start
+2024-05-13 20:46:53,323 DEBUG   HandlerThread:1244775 [system_info.py:__init__():26] System info init
+2024-05-13 20:46:53,323 DEBUG   HandlerThread:1244775 [system_info.py:__init__():41] System info init done
+2024-05-13 20:46:53,323 INFO    HandlerThread:1244775 [system_monitor.py:start():194] Starting system monitor
+2024-05-13 20:46:53,323 INFO    SystemMonitor:1244775 [system_monitor.py:_start():158] Starting system asset monitoring threads
+2024-05-13 20:46:53,324 INFO    HandlerThread:1244775 [system_monitor.py:probe():214] Collecting system info
+2024-05-13 20:46:53,324 INFO    SystemMonitor:1244775 [interfaces.py:start():188] Started cpu monitoring
+2024-05-13 20:46:53,325 INFO    SystemMonitor:1244775 [interfaces.py:start():188] Started disk monitoring
+2024-05-13 20:46:53,325 INFO    SystemMonitor:1244775 [interfaces.py:start():188] Started gpu monitoring
+2024-05-13 20:46:53,327 INFO    SystemMonitor:1244775 [interfaces.py:start():188] Started memory monitoring
+2024-05-13 20:46:53,329 INFO    SystemMonitor:1244775 [interfaces.py:start():188] Started network monitoring
+2024-05-13 20:46:53,365 DEBUG   HandlerThread:1244775 [system_info.py:probe():150] Probing system
+2024-05-13 20:46:53,366 DEBUG   HandlerThread:1244775 [system_info.py:_probe_git():135] Probing git
+2024-05-13 20:46:53,372 DEBUG   HandlerThread:1244775 [system_info.py:_probe_git():143] Probing git done
+2024-05-13 20:46:53,372 DEBUG   HandlerThread:1244775 [system_info.py:probe():198] Probing system done
+2024-05-13 20:46:53,372 DEBUG   HandlerThread:1244775 [system_monitor.py:probe():223] {'os': 'Linux-5.4.0-166-generic-x86_64-with-glibc2.31', 'python': '3.11.9', 'heartbeatAt': '2024-05-13T18:46:53.365083', 'startedAt': '2024-05-13T18:46:52.816759', 'docker': None, 'cuda': None, 'args': ('finetuning_concatenated_config.json',), 'state': 'running', 'program': '/raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/run_parler_tts_training.py', 'codePathLocal': 'run_parler_tts_training.py', 'codePath': 'run_parler_tts_training.py', 'git': {'remote': 'https://huggingface.co/sanchit-gandhi/parler-tts-mini-v0.1-expresso-concatenated-combined', 'commit': '50ba4323d7b8bb052629aa1b88283b9df081a821'}, 'email': '[email protected]', 'root': '/raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined', 'host': 'hf-dgx-01', 'username': 'sanchit', 'executable': '/home/sanchit/miniconda3/envs/venv/bin/python', 'cpu_count': 64, 'cpu_count_logical': 128, 'cpu_freq': {'current': 2257.736234375, 'min': 1500.0, 'max': 2250.0}, 'cpu_freq_per_core': [{'current': 1795.281, 'min': 1500.0, 'max': 2250.0}, {'current': 1794.292, 'min': 1500.0, 'max': 2250.0}, {'current': 1795.78, 'min': 1500.0, 'max': 2250.0}, {'current': 1792.55, 'min': 1500.0, 'max': 2250.0}, {'current': 1742.094, 'min': 1500.0, 'max': 2250.0}, {'current': 3026.54, 'min': 1500.0, 'max': 2250.0}, {'current': 1786.214, 'min': 1500.0, 'max': 2250.0}, {'current': 1742.547, 'min': 1500.0, 'max': 2250.0}, {'current': 1728.916, 'min': 1500.0, 'max': 2250.0}, {'current': 1734.023, 'min': 1500.0, 'max': 2250.0}, {'current': 3195.219, 'min': 1500.0, 'max': 2250.0}, {'current': 1733.722, 'min': 1500.0, 'max': 2250.0}, {'current': 3341.911, 'min': 1500.0, 'max': 2250.0}, {'current': 3325.608, 'min': 1500.0, 'max': 2250.0}, {'current': 3233.258, 'min': 1500.0, 'max': 2250.0}, {'current': 1669.847, 'min': 1500.0, 'max': 2250.0}, {'current': 1791.948, 'min': 1500.0, 'max': 2250.0}, {'current': 1796.794, 'min': 1500.0, 'max': 2250.0}, {'current': 1791.49, 'min': 1500.0, 'max': 2250.0}, {'current': 1793.945, 'min': 1500.0, 'max': 2250.0}, {'current': 3342.943, 'min': 1500.0, 'max': 2250.0}, {'current': 1669.791, 'min': 1500.0, 'max': 2250.0}, {'current': 1669.593, 'min': 1500.0, 'max': 2250.0}, {'current': 1694.312, 'min': 1500.0, 'max': 2250.0}, {'current': 1873.727, 'min': 1500.0, 'max': 2250.0}, {'current': 1724.813, 'min': 1500.0, 'max': 2250.0}, {'current': 2354.471, 'min': 1500.0, 'max': 2250.0}, {'current': 1718.662, 'min': 1500.0, 'max': 2250.0}, {'current': 1670.588, 'min': 1500.0, 'max': 2250.0}, {'current': 1665.577, 'min': 1500.0, 'max': 2250.0}, {'current': 1616.671, 'min': 1500.0, 'max': 2250.0}, {'current': 2080.81, 'min': 1500.0, 'max': 2250.0}, {'current': 1670.666, 'min': 1500.0, 'max': 2250.0}, {'current': 1652.559, 'min': 1500.0, 'max': 2250.0}, {'current': 3323.654, 'min': 1500.0, 'max': 2250.0}, {'current': 1671.311, 'min': 1500.0, 'max': 2250.0}, {'current': 1726.286, 'min': 1500.0, 'max': 2250.0}, {'current': 1670.365, 'min': 1500.0, 'max': 2250.0}, {'current': 3320.57, 'min': 1500.0, 'max': 2250.0}, {'current': 1669.941, 'min': 1500.0, 'max': 2250.0}, {'current': 1791.021, 'min': 1500.0, 'max': 2250.0}, {'current': 1796.246, 'min': 1500.0, 'max': 2250.0}, {'current': 1793.946, 'min': 1500.0, 'max': 2250.0}, {'current': 1794.848, 'min': 1500.0, 'max': 2250.0}, {'current': 3339.327, 'min': 1500.0, 'max': 2250.0}, {'current': 3344.315, 'min': 1500.0, 'max': 2250.0}, {'current': 3338.901, 'min': 1500.0, 'max': 2250.0}, {'current': 1668.541, 'min': 1500.0, 'max': 2250.0}, {'current': 1794.526, 'min': 1500.0, 'max': 2250.0}, {'current': 1792.886, 'min': 1500.0, 'max': 2250.0}, {'current': 1796.844, 'min': 1500.0, 'max': 2250.0}, {'current': 1793.81, 'min': 1500.0, 'max': 2250.0}, {'current': 1724.861, 'min': 1500.0, 'max': 2250.0}, {'current': 2294.458, 'min': 1500.0, 'max': 2250.0}, {'current': 1720.835, 'min': 1500.0, 'max': 2250.0}, {'current': 1720.155, 'min': 1500.0, 'max': 2250.0}, {'current': 1668.96, 'min': 1500.0, 'max': 2250.0}, {'current': 1976.5, 'min': 1500.0, 'max': 2250.0}, {'current': 2241.578, 'min': 1500.0, 'max': 2250.0}, {'current': 1671.964, 'min': 1500.0, 'max': 2250.0}, {'current': 3319.623, 'min': 1500.0, 'max': 2250.0}, {'current': 1670.777, 'min': 1500.0, 'max': 2250.0}, {'current': 1670.389, 'min': 1500.0, 'max': 2250.0}, {'current': 1669.629, 'min': 1500.0, 'max': 2250.0}, {'current': 1794.19, 'min': 1500.0, 'max': 2250.0}, {'current': 1794.138, 'min': 1500.0, 'max': 2250.0}, {'current': 1796.317, 'min': 1500.0, 'max': 2250.0}, {'current': 1792.821, 'min': 1500.0, 'max': 2250.0}, {'current': 1794.716, 'min': 1500.0, 'max': 2250.0}, {'current': 1793.624, 'min': 1500.0, 'max': 2250.0}, {'current': 1796.346, 'min': 1500.0, 'max': 2250.0}, {'current': 1793.897, 'min': 1500.0, 'max': 2250.0}, {'current': 1735.424, 'min': 1500.0, 'max': 2250.0}, {'current': 1738.64, 'min': 1500.0, 'max': 2250.0}, {'current': 1979.998, 'min': 1500.0, 'max': 2250.0}, {'current': 1737.286, 'min': 1500.0, 'max': 2250.0}, {'current': 3313.748, 'min': 1500.0, 'max': 2250.0}, {'current': 3337.223, 'min': 1500.0, 'max': 2250.0}, {'current': 1671.416, 'min': 1500.0, 'max': 2250.0}, {'current': 1670.005, 'min': 1500.0, 'max': 2250.0}, {'current': 1794.276, 'min': 1500.0, 'max': 2250.0}, {'current': 1738.22, 'min': 1500.0, 'max': 2250.0}, {'current': 1742.737, 'min': 1500.0, 'max': 2250.0}, {'current': 1770.535, 'min': 1500.0, 'max': 2250.0}, {'current': 3320.252, 'min': 1500.0, 'max': 2250.0}, {'current': 1671.037, 'min': 1500.0, 'max': 2250.0}, {'current': 1669.549, 'min': 1500.0, 'max': 2250.0}, {'current': 1670.948, 'min': 1500.0, 'max': 2250.0}, {'current': 2843.391, 'min': 1500.0, 'max': 2250.0}, {'current': 2348.589, 'min': 1500.0, 'max': 2250.0}, {'current': 3287.915, 'min': 1500.0, 'max': 2250.0}, {'current': 2340.192, 'min': 1500.0, 'max': 2250.0}, {'current': 2426.358, 'min': 1500.0, 'max': 2250.0}, {'current': 2415.833, 'min': 1500.0, 'max': 2250.0}, {'current': 2419.416, 'min': 1500.0, 'max': 2250.0}, {'current': 2277.433, 'min': 1500.0, 'max': 2250.0}, {'current': 2365.562, 'min': 1500.0, 'max': 2250.0}, {'current': 2400.6, 'min': 1500.0, 'max': 2250.0}, {'current': 2075.143, 'min': 1500.0, 'max': 2250.0}, {'current': 2382.295, 'min': 1500.0, 'max': 2250.0}, {'current': 3066.339, 'min': 1500.0, 'max': 2250.0}, {'current': 2466.631, 'min': 1500.0, 'max': 2250.0}, {'current': 3100.81, 'min': 1500.0, 'max': 2250.0}, {'current': 2421.93, 'min': 1500.0, 'max': 2250.0}, {'current': 3233.829, 'min': 1500.0, 'max': 2250.0}, {'current': 2234.583, 'min': 1500.0, 'max': 2250.0}, {'current': 2452.089, 'min': 1500.0, 'max': 2250.0}, {'current': 2975.985, 'min': 1500.0, 'max': 2250.0}, {'current': 3301.512, 'min': 1500.0, 'max': 2250.0}, {'current': 3336.905, 'min': 1500.0, 'max': 2250.0}, {'current': 2984.87, 'min': 1500.0, 'max': 2250.0}, {'current': 2384.306, 'min': 1500.0, 'max': 2250.0}, {'current': 2965.197, 'min': 1500.0, 'max': 2250.0}, {'current': 1929.067, 'min': 1500.0, 'max': 2250.0}, {'current': 1986.731, 'min': 1500.0, 'max': 2250.0}, {'current': 1999.412, 'min': 1500.0, 'max': 2250.0}, {'current': 2477.541, 'min': 1500.0, 'max': 2250.0}, {'current': 3111.851, 'min': 1500.0, 'max': 2250.0}, {'current': 2009.907, 'min': 1500.0, 'max': 2250.0}, {'current': 1993.784, 'min': 1500.0, 'max': 2250.0}, {'current': 2144.459, 'min': 1500.0, 'max': 2250.0}, {'current': 3337.426, 'min': 1500.0, 'max': 2250.0}, {'current': 3320.114, 'min': 1500.0, 'max': 2250.0}, {'current': 2169.719, 'min': 1500.0, 'max': 2250.0}, {'current': 3308.644, 'min': 1500.0, 'max': 2250.0}, {'current': 2111.633, 'min': 1500.0, 'max': 2250.0}, {'current': 2123.71, 'min': 1500.0, 'max': 2250.0}, {'current': 2153.49, 'min': 1500.0, 'max': 2250.0}], 'disk': {'/': {'total': 1757.8785285949707, 'used': 1663.5005989074707}}, 'gpu': 'NVIDIA A100-SXM4-80GB', 'gpu_count': 5, 'gpu_devices': [{'name': 'NVIDIA A100-SXM4-80GB', 'memory_total': 85899345920}, {'name': 'NVIDIA A100-SXM4-80GB', 'memory_total': 85899345920}, {'name': 'NVIDIA A100-SXM4-80GB', 'memory_total': 85899345920}, {'name': 'NVIDIA DGX Display', 'memory_total': 4294967296}, {'name': 'NVIDIA A100-SXM4-80GB', 'memory_total': 85899345920}], 'memory': {'total': 503.5396919250488}}
+2024-05-13 20:46:53,372 INFO    HandlerThread:1244775 [system_monitor.py:probe():224] Finished collecting system info
+2024-05-13 20:46:53,372 INFO    HandlerThread:1244775 [system_monitor.py:probe():227] Publishing system info
+2024-05-13 20:46:53,372 DEBUG   HandlerThread:1244775 [system_info.py:_save_conda():207] Saving list of conda packages installed into the current environment
+2024-05-13 20:46:53,387 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:46:53,400 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:46:54,224 INFO    Thread-12 :1244775 [dir_watcher.py:_on_file_created():271] file/dir created: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/conda-environment.yaml
+2024-05-13 20:46:55,418 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:46:55,429 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:46:55,741 DEBUG   HandlerThread:1244775 [system_info.py:_save_conda():222] Saving conda packages done
+2024-05-13 20:46:55,742 INFO    HandlerThread:1244775 [system_monitor.py:probe():229] Finished publishing system info
+2024-05-13 20:46:55,750 DEBUG   SenderThread:1244775 [sender.py:send():378] send: files
+2024-05-13 20:46:55,750 INFO    SenderThread:1244775 [sender.py:_save_file():1389] saving file wandb-metadata.json with policy now
+2024-05-13 20:46:55,863 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: python_packages
+2024-05-13 20:46:55,863 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:46:55,863 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: python_packages
+2024-05-13 20:46:55,865 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:46:56,093 DEBUG   SenderThread:1244775 [sender.py:send():378] send: telemetry
+2024-05-13 20:46:56,093 DEBUG   SenderThread:1244775 [sender.py:send():378] send: config
+2024-05-13 20:46:56,224 INFO    Thread-12 :1244775 [dir_watcher.py:_on_file_modified():288] file/dir modified: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/conda-environment.yaml
+2024-05-13 20:46:56,224 INFO    Thread-12 :1244775 [dir_watcher.py:_on_file_created():271] file/dir created: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/wandb-metadata.json
+2024-05-13 20:46:56,224 INFO    Thread-12 :1244775 [dir_watcher.py:_on_file_created():271] file/dir created: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/requirements.txt
+2024-05-13 20:46:56,224 INFO    Thread-12 :1244775 [dir_watcher.py:_on_file_created():271] file/dir created: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/output.log
+2024-05-13 20:46:56,261 INFO    wandb-upload_0:1244775 [upload_job.py:push():130] Uploaded file /tmp/tmpewnm1an9wandb/l0duo91p-wandb-metadata.json
+2024-05-13 20:46:58,225 INFO    Thread-12 :1244775 [dir_watcher.py:_on_file_modified():288] file/dir modified: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/output.log
+2024-05-13 20:46:58,331 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:46:58,343 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:46:58,386 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:47:00,362 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:00,375 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:03,387 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:47:03,431 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:03,443 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:05,466 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:05,478 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:08,388 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:47:08,679 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:08,692 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:10,713 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:10,724 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:10,863 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:47:10,863 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:47:12,746 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:12,757 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:14,094 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:47:15,627 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:15,637 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:17,658 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:17,668 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:19,096 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:47:20,779 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:20,799 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:22,817 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:22,830 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:24,099 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:47:24,858 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:24,870 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:25,233 INFO    Thread-12 :1244775 [dir_watcher.py:_on_file_modified():288] file/dir modified: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/config.yaml
+2024-05-13 20:47:25,863 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:47:25,863 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:47:27,736 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:27,747 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:30,092 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:47:31,414 ERROR   gpu       :1244775 [interfaces.py:aggregate():159] Failed to serialize metric: division by zero
+2024-05-13 20:47:31,434 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:31,462 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:34,494 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:34,518 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:35,093 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:47:36,569 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:36,599 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:38,635 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:38,658 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:40,093 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:47:40,863 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:47:40,864 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:47:41,560 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:41,584 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:43,631 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:43,652 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:46,084 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:47:46,570 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:46,604 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:48,647 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:48,659 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:51,084 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:47:51,664 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:51,686 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:53,329 DEBUG   SystemMonitor:1244775 [system_monitor.py:_start():172] Starting system metrics aggregation loop
+2024-05-13 20:47:53,333 DEBUG   SenderThread:1244775 [sender.py:send():378] send: stats
+2024-05-13 20:47:53,709 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:53,724 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:55,863 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:47:55,864 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:47:56,593 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:56,607 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:57,080 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:47:58,627 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:47:58,641 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:01,653 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:01,664 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:02,081 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:48:03,684 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:03,696 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:06,665 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:06,679 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:07,082 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:48:10,344 ERROR   gpu       :1244775 [interfaces.py:aggregate():159] Failed to serialize metric: division by zero
+2024-05-13 20:48:10,366 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:10,380 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:10,863 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:48:10,864 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:48:13,048 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:48:13,506 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:13,529 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:15,558 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:15,586 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:18,050 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:48:18,552 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:18,572 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:20,626 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:20,644 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:23,050 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:48:23,336 DEBUG   SenderThread:1244775 [sender.py:send():378] send: stats
+2024-05-13 20:48:23,683 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:23,707 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:25,750 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:25,769 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:25,863 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:48:25,864 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:48:28,681 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:28,701 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:29,005 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:48:30,725 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:30,747 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:33,782 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:33,801 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:34,006 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:48:35,835 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:35,858 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:38,877 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:38,904 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:39,007 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:48:40,863 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:48:40,864 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:48:40,932 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:40,946 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:42,969 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:42,980 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:44,036 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:48:45,801 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:45,831 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:49,037 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:48:49,512 ERROR   gpu       :1244775 [interfaces.py:aggregate():159] Failed to serialize metric: division by zero
+2024-05-13 20:48:49,550 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:49,567 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:52,479 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:52,494 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:53,338 DEBUG   SenderThread:1244775 [sender.py:send():378] send: stats
+2024-05-13 20:48:54,339 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:48:54,518 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:54,530 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:55,864 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:48:55,864 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:48:57,576 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:57,589 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:59,615 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:48:59,628 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:00,007 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:49:02,716 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:02,730 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:04,748 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:04,763 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:05,008 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:49:07,719 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:07,741 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:09,766 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:09,776 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:10,009 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:49:10,864 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:49:10,864 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:49:12,861 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:12,875 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:14,916 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:14,934 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:15,085 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:49:17,938 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:17,957 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:20,002 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:20,012 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:20,086 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:49:22,621 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:22,641 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:23,339 DEBUG   SenderThread:1244775 [sender.py:send():378] send: stats
+2024-05-13 20:49:24,675 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:24,686 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:25,341 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:49:25,864 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:49:25,865 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:49:28,370 ERROR   gpu       :1244775 [interfaces.py:aggregate():159] Failed to serialize metric: division by zero
+2024-05-13 20:49:28,612 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:28,633 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:30,656 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:30,678 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:31,038 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:49:33,121 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:33,150 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:35,177 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:35,217 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:36,039 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:49:37,932 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:37,967 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:40,002 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:40,032 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:40,864 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:49:40,864 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:49:42,037 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:49:42,295 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:42,620 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:44,641 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:44,688 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:47,037 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:49:47,242 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:47,286 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:49,336 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:49,362 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:51,898 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:51,927 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:52,038 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:49:53,343 DEBUG   SenderThread:1244775 [sender.py:send():378] send: stats
+2024-05-13 20:49:54,378 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:54,396 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:55,864 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:49:55,864 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:49:56,417 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:56,465 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:58,028 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:49:58,997 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:49:59,011 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:01,057 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:01,097 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:03,029 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:50:05,200 ERROR   gpu       :1244775 [interfaces.py:aggregate():159] Failed to serialize metric: division by zero
+2024-05-13 20:50:05,253 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:05,272 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:07,311 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:07,330 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:08,029 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:50:09,887 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:09,918 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:10,864 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:50:10,864 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:50:11,934 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:11,942 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:13,047 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:50:14,680 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:14,697 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:16,719 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:16,738 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:18,047 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:50:19,232 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:19,254 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:21,272 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:21,282 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:23,048 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:50:23,346 DEBUG   SenderThread:1244775 [sender.py:send():378] send: stats
+2024-05-13 20:50:23,802 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:23,817 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:25,839 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:25,850 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:25,864 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: stop_status
+2024-05-13 20:50:25,865 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: stop_status
+2024-05-13 20:50:28,061 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:50:28,256 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:28,265 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:30,284 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:30,296 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:32,293 INFO    Thread-12 :1244775 [dir_watcher.py:_on_file_modified():288] file/dir modified: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/output.log
+2024-05-13 20:50:32,787 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:32,796 ERROR   gpu       :1244775 [interfaces.py:monitor():142] Failed to sample metric: Not Supported
+2024-05-13 20:50:33,152 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:50:33,158 DEBUG   SenderThread:1244775 [sender.py:send():378] send: exit
+2024-05-13 20:50:33,158 INFO    SenderThread:1244775 [sender.py:send_exit():585] handling exit code: 1
+2024-05-13 20:50:33,159 INFO    SenderThread:1244775 [sender.py:send_exit():587] handling runtime: 219
+2024-05-13 20:50:33,159 INFO    SenderThread:1244775 [sender.py:_save_file():1389] saving file wandb-summary.json with policy end
+2024-05-13 20:50:33,159 INFO    SenderThread:1244775 [sender.py:send_exit():593] send defer
+2024-05-13 20:50:33,159 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:33,159 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 0
+2024-05-13 20:50:33,159 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:33,160 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 0
+2024-05-13 20:50:33,160 INFO    SenderThread:1244775 [sender.py:transition_state():613] send defer: 1
+2024-05-13 20:50:33,160 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:33,160 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 1
+2024-05-13 20:50:33,160 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:33,160 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 1
+2024-05-13 20:50:33,160 INFO    SenderThread:1244775 [sender.py:transition_state():613] send defer: 2
+2024-05-13 20:50:33,160 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:33,160 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 2
+2024-05-13 20:50:33,160 INFO    HandlerThread:1244775 [system_monitor.py:finish():203] Stopping system monitor
+2024-05-13 20:50:33,161 DEBUG   SystemMonitor:1244775 [system_monitor.py:_start():179] Finished system metrics aggregation loop
+2024-05-13 20:50:33,161 DEBUG   SystemMonitor:1244775 [system_monitor.py:_start():183] Publishing last batch of metrics
+2024-05-13 20:50:33,161 INFO    HandlerThread:1244775 [interfaces.py:finish():200] Joined cpu monitor
+2024-05-13 20:50:33,164 INFO    HandlerThread:1244775 [interfaces.py:finish():200] Joined disk monitor
+2024-05-13 20:50:33,293 INFO    Thread-12 :1244775 [dir_watcher.py:_on_file_created():271] file/dir created: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/wandb-summary.json
+2024-05-13 20:50:34,293 INFO    Thread-12 :1244775 [dir_watcher.py:_on_file_modified():288] file/dir modified: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/output.log
+2024-05-13 20:50:34,797 ERROR   gpu       :1244775 [interfaces.py:aggregate():159] Failed to serialize metric: division by zero
+2024-05-13 20:50:34,797 INFO    HandlerThread:1244775 [interfaces.py:finish():200] Joined gpu monitor
+2024-05-13 20:50:34,797 INFO    HandlerThread:1244775 [interfaces.py:finish():200] Joined memory monitor
+2024-05-13 20:50:34,797 INFO    HandlerThread:1244775 [interfaces.py:finish():200] Joined network monitor
+2024-05-13 20:50:34,798 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: poll_exit
+2024-05-13 20:50:34,799 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:34,799 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 2
+2024-05-13 20:50:34,799 INFO    SenderThread:1244775 [sender.py:transition_state():613] send defer: 3
+2024-05-13 20:50:34,800 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:34,800 DEBUG   SenderThread:1244775 [sender.py:send():378] send: stats
+2024-05-13 20:50:34,800 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 3
+2024-05-13 20:50:34,800 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: poll_exit
+2024-05-13 20:50:34,801 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:34,801 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 3
+2024-05-13 20:50:34,801 INFO    SenderThread:1244775 [sender.py:transition_state():613] send defer: 4
+2024-05-13 20:50:34,801 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:34,801 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 4
+2024-05-13 20:50:34,801 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:34,801 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 4
+2024-05-13 20:50:34,801 INFO    SenderThread:1244775 [sender.py:transition_state():613] send defer: 5
+2024-05-13 20:50:34,802 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:34,802 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 5
+2024-05-13 20:50:34,802 DEBUG   SenderThread:1244775 [sender.py:send():378] send: summary
+2024-05-13 20:50:34,802 INFO    SenderThread:1244775 [sender.py:_save_file():1389] saving file wandb-summary.json with policy end
+2024-05-13 20:50:34,802 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:34,802 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 5
+2024-05-13 20:50:34,802 INFO    SenderThread:1244775 [sender.py:transition_state():613] send defer: 6
+2024-05-13 20:50:34,802 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:34,802 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 6
+2024-05-13 20:50:34,803 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:34,803 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 6
+2024-05-13 20:50:34,803 INFO    SenderThread:1244775 [sender.py:transition_state():613] send defer: 7
+2024-05-13 20:50:34,803 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: status_report
+2024-05-13 20:50:34,803 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:34,803 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 7
+2024-05-13 20:50:34,803 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:34,803 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 7
+2024-05-13 20:50:35,159 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: poll_exit
+2024-05-13 20:50:35,294 INFO    Thread-12 :1244775 [dir_watcher.py:_on_file_modified():288] file/dir modified: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/wandb-summary.json
+2024-05-13 20:50:38,152 INFO    SenderThread:1244775 [sender.py:transition_state():613] send defer: 8
+2024-05-13 20:50:38,152 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: poll_exit
+2024-05-13 20:50:38,152 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:38,153 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 8
+2024-05-13 20:50:38,153 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:38,153 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 8
+2024-05-13 20:50:38,153 INFO    SenderThread:1244775 [job_builder.py:build():432] Attempting to build job artifact
+2024-05-13 20:50:38,153 INFO    SenderThread:1244775 [job_builder.py:_get_source_type():565] is repo sourced job
+2024-05-13 20:50:38,160 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: poll_exit
+2024-05-13 20:50:38,179 INFO    SenderThread:1244775 [job_builder.py:build():541] adding wandb-job metadata file
+2024-05-13 20:50:38,181 INFO    SenderThread:1244775 [sender.py:transition_state():613] send defer: 9
+2024-05-13 20:50:38,181 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: poll_exit
+2024-05-13 20:50:38,181 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:38,182 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 9
+2024-05-13 20:50:38,182 DEBUG   SenderThread:1244775 [sender.py:send():378] send: artifact
+2024-05-13 20:50:38,294 INFO    Thread-12 :1244775 [dir_watcher.py:_on_file_modified():288] file/dir modified: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/output.log
+2024-05-13 20:50:39,160 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: poll_exit
+2024-05-13 20:50:39,209 INFO    wandb-upload_0:1244775 [upload_job.py:push():88] Uploaded file /home/sanchit/.local/share/wandb/artifacts/staging/tmp34vs1_ku
+2024-05-13 20:50:39,238 INFO    wandb-upload_1:1244775 [upload_job.py:push():88] Uploaded file /tmp/tmp_ne7l6g3/wandb-job.json
+2024-05-13 20:50:40,085 INFO    SenderThread:1244775 [sender.py:send_artifact():1467] sent artifact job-https___huggingface.co_sanchit-gandhi_parler-tts-mini-v0.1-expresso-concatenated-combined_run_parler_tts_training.py - {'id': 'QXJ0aWZhY3Q6ODM0NzI5NzMx', 'state': 'PENDING', 'artifactSequence': {'id': 'QXJ0aWZhY3RDb2xsZWN0aW9uOjE3NDIzMTI3NQ==', 'latestArtifact': None}}
+2024-05-13 20:50:40,086 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:40,086 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 9
+2024-05-13 20:50:40,086 INFO    SenderThread:1244775 [dir_watcher.py:finish():358] shutting down directory watcher
+2024-05-13 20:50:40,295 INFO    SenderThread:1244775 [dir_watcher.py:finish():388] scan: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files
+2024-05-13 20:50:40,295 INFO    SenderThread:1244775 [dir_watcher.py:finish():402] scan save: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/conda-environment.yaml conda-environment.yaml
+2024-05-13 20:50:40,295 INFO    SenderThread:1244775 [dir_watcher.py:finish():402] scan save: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/wandb-summary.json wandb-summary.json
+2024-05-13 20:50:40,295 INFO    SenderThread:1244775 [dir_watcher.py:finish():402] scan save: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/output.log output.log
+2024-05-13 20:50:40,297 INFO    SenderThread:1244775 [dir_watcher.py:finish():402] scan save: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/requirements.txt requirements.txt
+2024-05-13 20:50:40,300 INFO    SenderThread:1244775 [dir_watcher.py:finish():402] scan save: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/config.yaml config.yaml
+2024-05-13 20:50:40,301 INFO    SenderThread:1244775 [dir_watcher.py:finish():402] scan save: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/wandb-metadata.json wandb-metadata.json
+2024-05-13 20:50:40,301 INFO    SenderThread:1244775 [sender.py:transition_state():613] send defer: 10
+2024-05-13 20:50:40,301 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: poll_exit
+2024-05-13 20:50:40,301 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:40,302 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 10
+2024-05-13 20:50:40,306 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:40,306 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 10
+2024-05-13 20:50:40,306 INFO    SenderThread:1244775 [file_pusher.py:finish():169] shutting down file pusher
+2024-05-13 20:50:40,632 INFO    wandb-upload_0:1244775 [upload_job.py:push():130] Uploaded file /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/conda-environment.yaml
+2024-05-13 20:50:40,639 INFO    wandb-upload_1:1244775 [upload_job.py:push():130] Uploaded file /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/output.log
+2024-05-13 20:50:40,758 INFO    wandb-upload_3:1244775 [upload_job.py:push():130] Uploaded file /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/requirements.txt
+2024-05-13 20:50:40,773 INFO    wandb-upload_2:1244775 [upload_job.py:push():130] Uploaded file /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/wandb-summary.json
+2024-05-13 20:50:40,918 INFO    wandb-upload_4:1244775 [upload_job.py:push():130] Uploaded file /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/files/config.yaml
+2024-05-13 20:50:41,118 INFO    Thread-11 (_thread_body):1244775 [sender.py:transition_state():613] send defer: 11
+2024-05-13 20:50:41,119 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:41,119 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 11
+2024-05-13 20:50:41,119 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:41,119 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 11
+2024-05-13 20:50:41,120 INFO    SenderThread:1244775 [file_pusher.py:join():175] waiting for file pusher
+2024-05-13 20:50:41,120 INFO    SenderThread:1244775 [sender.py:transition_state():613] send defer: 12
+2024-05-13 20:50:41,120 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:41,120 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 12
+2024-05-13 20:50:41,120 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:41,120 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 12
+2024-05-13 20:50:41,120 INFO    SenderThread:1244775 [file_stream.py:finish():601] file stream finish called
+2024-05-13 20:50:41,161 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: poll_exit
+2024-05-13 20:50:41,510 INFO    SenderThread:1244775 [file_stream.py:finish():605] file stream finish is done
+2024-05-13 20:50:41,510 INFO    SenderThread:1244775 [sender.py:transition_state():613] send defer: 13
+2024-05-13 20:50:41,510 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: poll_exit
+2024-05-13 20:50:41,510 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:41,510 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 13
+2024-05-13 20:50:41,510 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:41,510 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 13
+2024-05-13 20:50:41,510 INFO    SenderThread:1244775 [sender.py:transition_state():613] send defer: 14
+2024-05-13 20:50:41,511 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: defer
+2024-05-13 20:50:41,511 INFO    HandlerThread:1244775 [handler.py:handle_request_defer():184] handle defer: 14
+2024-05-13 20:50:41,511 DEBUG   SenderThread:1244775 [sender.py:send():378] send: final
+2024-05-13 20:50:41,511 DEBUG   SenderThread:1244775 [sender.py:send():378] send: footer
+2024-05-13 20:50:41,511 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: defer
+2024-05-13 20:50:41,511 INFO    SenderThread:1244775 [sender.py:send_request_defer():609] handle sender defer: 14
+2024-05-13 20:50:41,512 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: poll_exit
+2024-05-13 20:50:41,512 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: poll_exit
+2024-05-13 20:50:41,512 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: poll_exit
+2024-05-13 20:50:41,512 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: poll_exit
+2024-05-13 20:50:41,513 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: server_info
+2024-05-13 20:50:41,513 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: get_summary
+2024-05-13 20:50:41,513 DEBUG   SenderThread:1244775 [sender.py:send_request():405] send_request: server_info
+2024-05-13 20:50:41,515 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: sampled_history
+2024-05-13 20:50:41,515 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: internal_messages
+2024-05-13 20:50:41,651 INFO    MainThread:1244775 [wandb_run.py:_footer_history_summary_info():3994] rendering history
+2024-05-13 20:50:41,651 INFO    MainThread:1244775 [wandb_run.py:_footer_history_summary_info():4026] rendering summary
+2024-05-13 20:50:41,651 INFO    MainThread:1244775 [wandb_run.py:_footer_sync_info():3953] logging synced files
+2024-05-13 20:50:41,651 DEBUG   HandlerThread:1244775 [handler.py:handle_request():158] handle_request: shutdown
+2024-05-13 20:50:41,651 INFO    HandlerThread:1244775 [handler.py:finish():882] shutting down handler
+2024-05-13 20:50:42,513 INFO    WriterThread:1244775 [datastore.py:close():296] close: /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/run-m0g0ap7d.wandb
+2024-05-13 20:50:42,651 INFO    SenderThread:1244775 [sender.py:finish():1545] shutting down sender
+2024-05-13 20:50:42,651 INFO    SenderThread:1244775 [file_pusher.py:finish():169] shutting down file pusher
+2024-05-13 20:50:42,651 INFO    SenderThread:1244775 [file_pusher.py:join():175] waiting for file pusher

wandb/run-20240513_204652-m0g0ap7d/logs/debug.log

@@ -0,0 +1,29 @@
+2024-05-13 20:46:52,817 INFO    MainThread:1244577 [wandb_setup.py:_flush():76] Current SDK version is 0.17.0
+2024-05-13 20:46:52,817 INFO    MainThread:1244577 [wandb_setup.py:_flush():76] Configure stats pid to 1244577
+2024-05-13 20:46:52,817 INFO    MainThread:1244577 [wandb_setup.py:_flush():76] Loading settings from /home/sanchit/.config/wandb/settings
+2024-05-13 20:46:52,817 INFO    MainThread:1244577 [wandb_setup.py:_flush():76] Loading settings from /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/settings
+2024-05-13 20:46:52,817 INFO    MainThread:1244577 [wandb_setup.py:_flush():76] Loading settings from environment variables: {}
+2024-05-13 20:46:52,817 INFO    MainThread:1244577 [wandb_setup.py:_flush():76] Applying setup settings: {'_disable_service': False}
+2024-05-13 20:46:52,817 INFO    MainThread:1244577 [wandb_setup.py:_flush():76] Inferring run settings from compute environment: {'program_relpath': 'run_parler_tts_training.py', 'program_abspath': '/raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/run_parler_tts_training.py', 'program': '/raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/run_parler_tts_training.py'}
+2024-05-13 20:46:52,817 INFO    MainThread:1244577 [wandb_setup.py:_flush():76] Applying login settings: {}
+2024-05-13 20:46:52,817 INFO    MainThread:1244577 [wandb_init.py:_log_setup():520] Logging user logs to /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/logs/debug.log
+2024-05-13 20:46:52,818 INFO    MainThread:1244577 [wandb_init.py:_log_setup():521] Logging internal logs to /raid/sanchit/parler-tts-mini-v0.1-expresso-concatenated-combined/wandb/run-20240513_204652-m0g0ap7d/logs/debug-internal.log
+2024-05-13 20:46:52,818 INFO    MainThread:1244577 [wandb_init.py:init():560] calling init triggers
+2024-05-13 20:46:52,818 INFO    MainThread:1244577 [wandb_init.py:init():567] wandb.init called with sweep_config: {}
+config: {}
+2024-05-13 20:46:52,818 INFO    MainThread:1244577 [wandb_init.py:init():610] starting backend
+2024-05-13 20:46:52,818 INFO    MainThread:1244577 [wandb_init.py:init():614] setting up manager
+2024-05-13 20:46:52,821 INFO    MainThread:1244577 [backend.py:_multiprocessing_setup():105] multiprocessing start_methods=fork,spawn,forkserver, using: spawn
+2024-05-13 20:46:52,822 INFO    MainThread:1244577 [wandb_init.py:init():622] backend started and connected
+2024-05-13 20:46:52,823 INFO    MainThread:1244577 [wandb_init.py:init():711] updated telemetry
+2024-05-13 20:46:52,826 INFO    MainThread:1244577 [wandb_init.py:init():744] communicating run to backend with 90.0 second timeout
+2024-05-13 20:46:53,228 INFO    MainThread:1244577 [wandb_run.py:_on_init():2396] communicating current version
+2024-05-13 20:46:53,287 INFO    MainThread:1244577 [wandb_run.py:_on_init():2405] got version response 
+2024-05-13 20:46:53,287 INFO    MainThread:1244577 [wandb_init.py:init():795] starting run threads in backend
+2024-05-13 20:46:55,863 INFO    MainThread:1244577 [wandb_run.py:_console_start():2374] atexit reg
+2024-05-13 20:46:55,863 INFO    MainThread:1244577 [wandb_run.py:_redirect():2229] redirect: wrap_raw
+2024-05-13 20:46:55,863 INFO    MainThread:1244577 [wandb_run.py:_redirect():2294] Wrapping output streams.
+2024-05-13 20:46:55,863 INFO    MainThread:1244577 [wandb_run.py:_redirect():2319] Redirects installed.
+2024-05-13 20:46:55,864 INFO    MainThread:1244577 [wandb_init.py:init():838] run started, returning control to user process
+2024-05-13 20:46:55,864 INFO    MainThread:1244577 [wandb_run.py:_config_callback():1376] config_cb None None {'learning_rate': 8e-05, 'model_name_or_path': 'parler-tts/parler_tts_mini_v0.1', 'num_train_epochs': 8, 'gradient_accumulation_steps': 8, 'per_device_train_batch_size': 16, 'global_batch_size': 16, 'mixed_precision': 'bf16', 'lr_scheduler_type': 'SchedulerType.COSINE', 'warmup_steps': 250, 'freeze_text_encoder': True, 'max_duration_in_seconds': 30.0, 'weight_decay': 0.01, 'adam_beta1': 0.9, 'adam_beta2': 0.99, 'temperature': 1.0}
+2024-05-13 20:50:42,653 WARNING MsgRouterThr:1244577 [router.py:message_loop():77] message_loop has been closed

wandb/run-20240513_205249-qaoje1x9/files/conda-environment.yaml

@@ -0,0 +1,248 @@
+name: venv
+channels:
+  - defaults
+dependencies:
+  - _libgcc_mutex=0.1=main
+  - _openmp_mutex=5.1=1_gnu
+  - bzip2=1.0.8=h5eee18b_6
+  - ca-certificates=2024.3.11=h06a4308_0
+  - ld_impl_linux-64=2.38=h1181459_1
+  - libffi=3.4.4=h6a678d5_1
+  - libgcc-ng=11.2.0=h1234567_1
+  - libgomp=11.2.0=h1234567_1
+  - libstdcxx-ng=11.2.0=h1234567_1
+  - libuuid=1.41.5=h5eee18b_0
+  - ncurses=6.4=h6a678d5_0
+  - openssl=3.0.13=h7f8727e_1
+  - pip=24.0=py311h06a4308_0
+  - python=3.11.9=h955ad1f_0
+  - readline=8.2=h5eee18b_0
+  - setuptools=69.5.1=py311h06a4308_0
+  - sqlite=3.45.3=h5eee18b_0
+  - tk=8.6.14=h39e8969_0
+  - wheel=0.43.0=py311h06a4308_0
+  - xz=5.4.6=h5eee18b_1
+  - zlib=1.2.13=h5eee18b_1
+  - pip:
+      - absl-py==2.1.0
+      - accelerate==0.30.0
+      - aiohttp==3.9.5
+      - aiosignal==1.3.1
+      - aniso8601==9.0.1
+      - annotated-types==0.6.0
+      - anyio==4.3.0
+      - argbind==0.3.7
+      - argon2-cffi==23.1.0
+      - argon2-cffi-bindings==21.2.0
+      - arrow==1.3.0
+      - asttokens==2.4.1
+      - async-lru==2.0.4
+      - attrs==23.2.0
+      - audioread==3.0.1
+      - babel==2.15.0
+      - beautifulsoup4==4.12.3
+      - bidict==0.23.1
+      - bitsandbytes==0.43.1
+      - bleach==6.1.0
+      - certifi==2024.2.2
+      - cffi==1.16.0
+      - charset-normalizer==3.3.2
+      - click==8.1.7
+      - coloredlogs==14.0
+      - comm==0.2.2
+      - contourpy==1.2.1
+      - cycler==0.12.1
+      - datasets==2.19.1
+      - debugpy==1.8.1
+      - decorator==5.1.1
+      - defusedxml==0.7.1
+      - descript-audio-codec==1.0.0
+      - descript-audiotools==0.7.2
+      - dill==0.3.8
+      - dnspython==2.3.0
+      - docker-pycreds==0.4.0
+      - docstring-parser==0.16
+      - editdistance==0.8.1
+      - einops==0.8.0
+      - et-xmlfile==1.1.0
+      - evaluate==0.4.2
+      - eventlet==0.36.1
+      - executing==2.0.1
+      - fastjsonschema==2.19.1
+      - ffmpy==0.3.2
+      - filelock==3.14.0
+      - fire==0.6.0
+      - flask==2.2.5
+      - flask-cors==4.0.1
+      - flask-restful==0.3.10
+      - flask-socketio==5.3.6
+      - flask-talisman==1.1.0
+      - flatten-dict==0.4.2
+      - fonttools==4.51.0
+      - fqdn==1.5.1
+      - frozenlist==1.4.1
+      - fsspec==2024.3.1
+      - future==1.0.0
+      - g2p==2.0.0
+      - gitdb==4.0.11
+      - gitpython==3.1.43
+      - greenlet==3.0.3
+      - grpcio==1.63.0
+      - h11==0.14.0
+      - httpcore==1.0.5
+      - httpx==0.27.0
+      - huggingface-hub==0.23.0
+      - humanfriendly==10.0
+      - idna==3.7
+      - importlib-resources==6.4.0
+      - ipdb==0.13.13
+      - ipykernel==6.29.4
+      - ipython==8.24.0
+      - isoduration==20.11.0
+      - itsdangerous==2.2.0
+      - jedi==0.19.1
+      - jinja2==3.1.4
+      - jiwer==3.0.4
+      - joblib==1.4.2
+      - json5==0.9.25
+      - jsonpointer==2.4
+      - jsonschema==4.22.0
+      - jsonschema-specifications==2023.12.1
+      - julius==0.2.7
+      - jupyter-client==8.6.1
+      - jupyter-core==5.7.2
+      - jupyter-events==0.10.0
+      - jupyter-lsp==2.2.5
+      - jupyter-server==2.14.0
+      - jupyter-server-terminals==0.5.3
+      - jupyterlab==4.2.0
+      - jupyterlab-pygments==0.3.0
+      - jupyterlab-server==2.27.1
+      - kiwisolver==1.4.5
+      - lazy-loader==0.4
+      - librosa==0.10.2
+      - llvmlite==0.42.0
+      - markdown==3.6
+      - markdown-it-py==3.0.0
+      - markdown2==2.4.13
+      - markupsafe==2.1.5
+      - matplotlib==3.8.4
+      - matplotlib-inline==0.1.7
+      - mdurl==0.1.2
+      - mistune==3.0.2
+      - mpmath==1.3.0
+      - msgpack==1.0.8
+      - multidict==6.0.5
+      - multiprocess==0.70.16
+      - munkres==1.1.4
+      - nbclient==0.10.0
+      - nbconvert==7.16.4
+      - nbformat==5.10.4
+      - nest-asyncio==1.6.0
+      - networkx==3.3
+      - notebook-shim==0.2.4
+      - numba==0.59.1
+      - numpy==1.26.4
+      - nvidia-cublas-cu12==12.1.3.1
+      - nvidia-cuda-cupti-cu12==12.1.105
+      - nvidia-cuda-nvrtc-cu12==12.1.105
+      - nvidia-cuda-runtime-cu12==12.1.105
+      - nvidia-cudnn-cu12==8.9.2.26
+      - nvidia-cufft-cu12==11.0.2.54
+      - nvidia-curand-cu12==10.3.2.106
+      - nvidia-cusolver-cu12==11.4.5.107
+      - nvidia-cusparse-cu12==12.1.0.106
+      - nvidia-nccl-cu12==2.20.5
+      - nvidia-nvjitlink-cu12==12.4.127
+      - nvidia-nvtx-cu12==12.1.105
+      - openpyxl==3.1.2
+      - overrides==7.7.0
+      - packaging==24.0
+      - pandas==2.2.2
+      - pandocfilters==1.5.1
+      - panphon==0.20.0
+      - parler-tts==0.1
+      - parso==0.8.4
+      - pexpect==4.9.0
+      - pillow==10.3.0
+      - platformdirs==4.2.1
+      - pooch==1.8.1
+      - prometheus-client==0.20.0
+      - prompt-toolkit==3.0.43
+      - protobuf==3.19.6
+      - psutil==5.9.8
+      - ptyprocess==0.7.0
+      - pure-eval==0.2.2
+      - pyarrow==16.0.0
+      - pyarrow-hotfix==0.6
+      - pycparser==2.22
+      - pydantic==2.7.1
+      - pydantic-core==2.18.2
+      - pygments==2.18.0
+      - pyloudnorm==0.1.1
+      - pyparsing==3.1.2
+      - pystoi==0.4.1
+      - python-dateutil==2.9.0.post0
+      - python-engineio==4.9.0
+      - python-json-logger==2.0.7
+      - python-socketio==5.11.2
+      - pytz==2024.1
+      - pyyaml==6.0.1
+      - pyzmq==26.0.3
+      - randomname==0.2.1
+      - rapidfuzz==3.9.0
+      - referencing==0.35.1
+      - regex==2024.4.28
+      - requests==2.31.0
+      - rfc3339-validator==0.1.4
+      - rfc3986-validator==0.1.1
+      - rich==13.7.1
+      - rpds-py==0.18.1
+      - safetensors==0.4.3
+      - scikit-learn==1.4.2
+      - scipy==1.13.0
+      - send2trash==1.8.3
+      - sentencepiece==0.2.0
+      - sentry-sdk==2.1.1
+      - setproctitle==1.3.3
+      - simple-websocket==1.0.0
+      - six==1.16.0
+      - smmap==5.0.1
+      - sniffio==1.3.1
+      - soundfile==0.12.1
+      - soupsieve==2.5
+      - soxr==0.3.7
+      - stack-data==0.6.3
+      - sympy==1.12
+      - tensorboard==2.16.2
+      - tensorboard-data-server==0.7.2
+      - termcolor==2.4.0
+      - terminado==0.18.1
+      - text-unidecode==1.3
+      - threadpoolctl==3.5.0
+      - tinycss2==1.3.0
+      - tokenizers==0.19.1
+      - torch==2.3.0
+      - torch-stoi==0.2.1
+      - torchaudio==2.3.0
+      - tornado==6.4
+      - tqdm==4.66.4
+      - traitlets==5.14.3
+      - transformers==4.41.0.dev0
+      - triton==2.3.0
+      - types-python-dateutil==2.9.0.20240316
+      - typing-extensions==4.11.0
+      - tzdata==2024.1
+      - unicodecsv==0.14.1
+      - uri-template==1.3.0
+      - urllib3==2.2.1
+      - wandb==0.17.0
+      - wcwidth==0.2.13
+      - webcolors==1.13
+      - webencodings==0.5.1
+      - websocket-client==1.8.0
+      - werkzeug==3.0.3
+      - wsproto==1.2.0
+      - xxhash==3.4.1
+      - yarl==1.9.4
+prefix: /home/sanchit/miniconda3/envs/venv

wandb/run-20240513_205249-qaoje1x9/files/config.yaml

@@ -0,0 +1,88 @@
+wandb_version: 1
+
+_wandb:
+  desc: null
+  value:
+    python_version: 3.11.9
+    cli_version: 0.17.0
+    framework: huggingface
+    huggingface_version: 4.41.0.dev0
+    is_jupyter_run: false
+    is_kaggle_kernel: false
+    start_time: 1715626369
+    t:
+      1:
+      - 1
+      - 5
+      - 11
+      - 49
+      - 51
+      - 53
+      - 55
+      - 71
+      - 100
+      2:
+      - 1
+      - 5
+      - 11
+      - 49
+      - 51
+      - 53
+      - 55
+      - 71
+      - 100
+      3:
+      - 2
+      - 23
+      - 61
+      4: 3.11.9
+      5: 0.17.0
+      6: 4.41.0.dev0
+      8:
+      - 5
+      13: linux-x86_64
+learning_rate:
+  desc: null
+  value: 8.0e-05
+model_name_or_path:
+  desc: null
+  value: parler-tts/parler_tts_mini_v0.1
+num_train_epochs:
+  desc: null
+  value: 8
+gradient_accumulation_steps:
+  desc: null
+  value: 8
+per_device_train_batch_size:
+  desc: null
+  value: 16
+global_batch_size:
+  desc: null
+  value: 16
+mixed_precision:
+  desc: null
+  value: bf16
+lr_scheduler_type:
+  desc: null
+  value: SchedulerType.COSINE
+warmup_steps:
+  desc: null
+  value: 250
+freeze_text_encoder:
+  desc: null
+  value: true
+max_duration_in_seconds:
+  desc: null
+  value: 30.0
+weight_decay:
+  desc: null
+  value: 0.01
+adam_beta1:
+  desc: null
+  value: 0.9
+adam_beta2:
+  desc: null
+  value: 0.99
+temperature:
+  desc: null
+  value: 1.0

wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech samples/eval_381_4899e0da4615e883ad13.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4899e0da4615e883ad137c4ebecee7d753f2512e69b2610ee121046fafa228a2
+size 1282092

wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech samples/eval_381_492597073098578f0605.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:492597073098578f0605b080944b218a8c03fd49715aa37e5b6b9f2bbdc975fe
+size 1282092

wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech samples/eval_381_67d9409a306e3614ec3f.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67d9409a306e3614ec3f84ea766046f8674e3cd4a8b13d1231dd6a356164f6d3
+size 1282092

wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech samples/eval_381_7c47ba927ac118ffaacc.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c47ba927ac118ffaacca70183d37077fd7a6d6fc2c649e51b7fb5137609cd8e
+size 1282092

wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech samples/eval_381_89ad32d31f3e70178cc1.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89ad32d31f3e70178cc19b1fb1c8ceda0377b104575def497602c02ba8bd1196
+size 1282092

wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech samples/eval_381_cd644667186ae0518a3c.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd644667186ae0518a3c08763c4aabf1e1f4d5754f3025b1e40779e8f068d489
+size 1282092

wandb/run-20240513_205249-qaoje1x9/files/media/audio/Speech samples/eval_381_f7405ef7b645b3265477.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f7405ef7b645b3265477b3e4e65bf1313dbc4445631e516872a054a694bb5e4a
+size 1282092