RuntimeError: CUDA out of memory

Sadam1195 · June 8, 2020, 5:59am

I have some custom dataset which is about 40 hours voice data.
Some utterances are larger (1 and half mint, there are many which are longer than 40 seconds) which I think is causing the issue but I need your comments on that and some are of very short time duration (1 sec, 2 sec, etc).
I tried to train it on different VM machines but on every machine I get the same error “CUDA out of memory”.
The maximum memory I had was 16 GB on Tesla V100 machine.
I used “use_phonemes”: true, “phoneme_language”: “it”.
I tried smaller batch sizes but it didn’t work either.
I am attaching my config.json and logs of few vm machines which failed to do the training.
I used CUDA_VISIBLE_DEVICES=“0,1,3,4,5,6,7” python distribute.py --config_path config.json after few steps it failed it gave me the error.

Please review these files and help me sort this out and also what is the best way to estimate the GPU memory required to train on a dataset, is there any way to calculate that?

I can’t upload the files “new user”

here is the output of my config.json

{
“model”: “Tacotron2”,
“run_name”: “ljspeech”,
“run_description”: “tacotron2”,

// AUDIO PARAMETERS
"audio":{
    // stft parameters
    "num_freq": 513,         // number of stft frequency levels. Size of the linear spectogram frame.
    "win_length": 1024,      // stft window length in ms.
    "hop_length": 256,       // stft window hop-lengh in ms.
    "frame_length_ms": null, // stft window length in ms.If null, 'win_length' is used.
    "frame_shift_ms": null,  // stft window hop-lengh in ms. If null, 'hop_length' is used.

    // Audio processing parameters
    "sample_rate": 22050,   // DATASET-RELATED: wav sample-rate. If different than the original data, it is resampled.
    "preemphasis": 0.0,     // pre-emphasis to reduce spec noise and make it more structured. If 0.0, no -pre-emphasis.
    "ref_level_db": 20,     // reference level db, theoretically 20db is the sound of air.

    // Silence trimming
    "do_trim_silence": true,// enable trimming of slience of audio as you load it. LJspeech (false), TWEB (false), Nancy (true)
    "trim_db": 60,          // threshold for timming silence. Set this according to your dataset.

    // Griffin-Lim
    "power": 1.5,           // value to sharpen wav signals after GL algorithm.
    "griffin_lim_iters": 60,// #griffin-lim iterations. 30-60 is a good range. Larger the value, slower the generation.

    // MelSpectrogram parameters
    "num_mels": 80,         // size of the mel spec frame.
    "mel_fmin": 0.0,        // minimum freq level for mel-spec. ~50 for male and ~95 for female voices. Tune for dataset!!
    "mel_fmax": 8000.0,     // maximum freq level for mel-spec. Tune for dataset!!

    // Normalization parameters
    "signal_norm": true,    // normalize spec values. Mean-Var normalization if 'stats_path' is defined otherwise range normalization defined by the other params.
    "min_level_db": -100,   // lower bound for normalization
    "symmetric_norm": true, // move normalization to range [-1, 1]
    "max_norm": 4.0,        // scale normalization to range [-max_norm, max_norm] or [0, max_norm]
    "clip_norm": true,      // clip normalized values into the range.
    "stats_path": null    // DO NOT USE WITH MULTI_SPEAKER MODEL. scaler stats file computed by 'compute_statistics.py'. If it is defined, mean-std based notmalization is used and other normalization params are ignored
},

// VOCABULARY PARAMETERS
// if custom character set is not defined,
// default set in symbols.py is used
// "characters":{
//     "pad": "_",
//     "eos": "~",
//     "bos": "^",
//     "characters": "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz!'(),-.:;? ",
//     "punctuations":"!'(),-.:;? ",
//     "phonemes":"iyɨʉɯuɪʏʊeøɘəɵɤoɛœɜɞʌɔæɐaɶɑɒᵻʘɓǀɗǃʄǂɠǁʛpbtdʈɖcɟkɡqɢʔɴŋɲɳnɱmʙrʀⱱɾɽɸβfvθðszʃʒʂʐçʝxɣχʁħʕhɦɬɮʋɹɻjɰlɭʎʟˈˌːˑʍwɥʜʢʡɕʑɺɧɚ˞ɫ"
// },

// DISTRIBUTED TRAINING
"distributed":{
    "backend": "nccl",
    "url": "tcp:\/\/localhost:54321"
},

"reinit_layers": [],    // give a list of layer names to restore from the given checkpoint. If not defined, it reloads all heuristically matching layers.

// TRAINING
"batch_size": 32,       // Batch size for training. Lower values than 32 might cause hard to learn attention. It is overwritten by 'gradual_training'.
"eval_batch_size":16,
"r": 7,                 // Number of decoder frames to predict per iteration. Set the initial values if gradual training is enabled.
"gradual_training": [[0, 7, 64], [1, 5, 64], [50000, 3, 32], [130000, 2, 32], [290000, 1, 32]], //set gradual training steps [first_step, r, batch_size]. If it is null, gradual training is disabled. For Tacotron, you might need to reduce the 'batch_size' as you proceeed.
"loss_masking": true,         // enable / disable loss masking against the sequence padding.
"ga_alpha": 10.0,        // weight for guided attention loss. If > 0, guided attention is enabled.

// VALIDATION
"run_eval": true,
"test_delay_epochs": 10,  //Until attention is aligned, testing only wastes computation time.
"test_sentences_file": null,  // set a file to load sentences to be used for testing. If it is null then we use default english sentences.

// OPTIMIZER
"noam_schedule": false,        // use noam warmup and lr schedule.
"grad_clip": 1.0,              // upper limit for gradients for clipping.
"epochs": 1000,                // total number of epochs to train.
"lr": 0.0001,                  // Initial learning rate. If Noam decay is active, maximum learning rate.
"wd": 0.000001,                // Weight decay weight.
"warmup_steps": 4000,          // Noam decay steps to increase the learning rate from 0 to "lr"
"seq_len_norm": false,         // Normalize eash sample loss with its length to alleviate imbalanced datasets. Use it if your dataset is small or has skewed distribution of sequence lengths.

// TACOTRON PRENET
"memory_size": -1,             // ONLY TACOTRON - size of the memory queue used fro storing last decoder predictions for auto-regression. If < 0, memory queue is disabled and decoder only uses the last prediction frame.
"prenet_type": "original",     // "original" or "bn".
"prenet_dropout": true,        // enable/disable dropout at prenet.

// ATTENTION
"attention_type": "original",  // 'original' or 'graves'
"attention_heads": 4,          // number of attention heads (only for 'graves')
"attention_norm": "sigmoid",   // softmax or sigmoid. Suggested to use softmax for Tacotron2 and sigmoid for Tacotron.
"windowing": false,            // Enables attention windowing. Used only in eval mode.
"use_forward_attn": false,     // if it uses forward attention. In general, it aligns faster.
"forward_attn_mask": false,    // Additional masking forcing monotonicity only in eval mode.
"transition_agent": false,     // enable/disable transition agent of forward attention.
"location_attn": true,         // enable_disable location sensitive attention. It is enabled for TACOTRON by default.
"bidirectional_decoder": false,  // use https://arxiv.org/abs/1907.09006. Use it, if attention does not work well with your dataset.

// STOPNET
"stopnet": true,               // Train stopnet predicting the end of synthesis.
"separate_stopnet": true,      // Train stopnet seperately if 'stopnet==true'. It prevents stopnet loss to influence the rest of the model. It causes a better model, but it trains SLOWER.

// TENSORBOARD and LOGGING
"print_step": 25,       // Number of steps to log traning on console.
"print_eval": false,     // If True, it prints intermediate loss values in evalulation.
"save_step": 10000,      // Number of training steps expected to save traninpg stats and checkpoints.
"checkpoint": true,     // If true, it saves checkpoints per "save_step"
"tb_model_param_stats": false,     // true, plots param stats per layer on tensorboard. Might be memory consuming, but good for debugging.

// DATA LOADING
"text_cleaner": "phoneme_cleaners",
"enable_eos_bos_chars": false, // enable/disable beginning of sentence and end of sentence chars.
"num_loader_workers": 4,        // number of training data loader processes. Don't set it too big. 4-8 are good values.
"num_val_loader_workers": 4,    // number of evaluation data loader processes.
"batch_group_size": 0,  //Number of batches to shuffle after bucketing.
"min_seq_len": 6,       // DATASET-RELATED: minimum text length to use in training
"max_seq_len": 153,     // DATASET-RELATED: maximum text length

// PATHS
"output_path": "/home/sadam/Projects/Italian_TTS/TTS/Models/LJSpeech/",

// PHONEMES
"phoneme_cache_path": "/home/sadam/Projects/Italian_TTS/TTS/mozilla_us_phonemes_3",  // phoneme computation is slow, therefore, it caches results in the given folder.
"use_phonemes": true,           // use phonemes instead of raw characters. It is suggested for better pronounciation.
"phoneme_language": "it",     // depending on your target language, pick one from  https://github.com/bootphon/phonemizer#languages

// MULTI-SPEAKER and GST
"use_speaker_embedding": false,     // use speaker embedding to enable multi-speaker learning.
"style_wav_for_test": null,          // path to style wav file to be used in TacotronGST inference.
"use_gst": false,       // TACOTRON ONLY: use global style tokens

// DATASETS
"datasets":   // List of datasets. They all merged and they get different speaker_ids.
    [
        {
            "name": "ljspeech",
            "path": "/home/sadam/Projects/Italian_TTS/LJSpeech-1.1/",
            "meta_file_train": "metadata.csv",
            "meta_file_val": null
        }
    ]

}

Here is the text from logs of different vm machines.

log_1_tesla_v_100

Using CUDA: True
Number of GPUs: 1
Git Hash: 8f8ba5e
Experiment folder: …/LJSpeech/ljspeech-June-07-2020_06+46PM-8f8ba5e
Setting up Audio Processor…
| > sample_rate:22050
| > num_mels:80
| > min_level_db:-100
| > frame_shift_ms:None
| > frame_length_ms:None
| > ref_level_db:20
| > num_freq:513
| > power:1.5
| > preemphasis:0.0
| > griffin_lim_iters:60
| > signal_norm:True
| > symmetric_norm:True
| > mel_fmin:0
| > mel_fmax:8000.0
| > max_norm:4.0
| > clip_norm:True
| > do_trim_silence:True
| > trim_db:60
| > do_sound_norm:False
| > stats_path:None
| > hop_length:256
| > win_length:1024
| > n_fft:1024
Using model: Tacotron2
| > Num output units : 513

Model has 28921234 parameters

EPOCH: 0/1000

Number of output frames: 7

DataLoader initialization
| > Use phonemes: True
| > phoneme language: it
| > Number of instances : 14820
| > Max length sequence: 1029
| > Min length sequence: 2
| > Avg length sequence: 113.39838056680162
| > Num. instances discarded by max-min (max=153, min=6) seq limits: 4153
| > Batch group size: 0.

TRAINING (2020-06-07 18:46:37)

–> STEP: 99/166 – GLOBAL_STEP: 100
| > decoder_loss: 2.77750 (5.47478)
| > postnet_loss: 2.61901 (5.85372)
| > stopnet_loss: 0.40204 (0.68065)
| > ga_loss: 0.03796 (0.10869)
| > loss: 5.43447
| > align_error: 0.98513 (0.96466)
| > avg_spec_len: 431.40625
| > avg_text_len: 77.59375
| > step_time: 0.51
| > loader_time: 0.01
| > lr: 0.00010
[WARNING] fount 1 utterances containing language switches on lines 2
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)

–> STEP: 124/166 – GLOBAL_STEP: 125
| > decoder_loss: 2.40896 (4.89162)
| > postnet_loss: 2.30191 (5.19040)
| > stopnet_loss: 0.33852 (0.61896)
| > ga_loss: 0.02859 (0.09361)
| > loss: 4.73946
| > align_error: 0.98797 (0.96909)
| > avg_spec_len: 599.25
| > avg_text_len: 103.390625
| > step_time: 0.70
| > loader_time: 19.06
| > lr: 0.00010
[WARNING] fount 1 utterances containing language switches on lines 1
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)
[WARNING] fount 1 utterances containing language switches on lines 2
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)
[WARNING] fount 1 utterances containing language switches on lines 4
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)

–> STEP: 149/166 – GLOBAL_STEP: 150
| > decoder_loss: 1.87681 (4.43319)
| > postnet_loss: 1.90342 (4.68345)
| > stopnet_loss: 0.31595 (0.57050)
| > ga_loss: 0.02243 (0.08223)
| > loss: 3.80266
| > align_error: 0.99110 (0.97259)
| > avg_spec_len: 727.28125
| > avg_text_len: 133.703125
| > step_time: 0.80
| > loader_time: 3.95
| > lr: 0.00010
[WARNING] fount 1 utterances containing language switches on lines 2
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)

EVALUATION

warning: audio amplitude out of range, auto clipped.
–> EVAL PERFORMANCE
| > avg_postnet_loss: 1.50400 (+0.00000)
| > avg_decoder_loss: 1.80960 (+0.00000)
| > avg_stopnet_loss: 0.60253 (+0.00000)
| > avg_align_error: 0.97442 (+0.00000)
| > avg_ga_loss: 0.06078 (+0.00000)

BEST MODEL : …/LJSpeech/ljspeech-June-07-2020_06+46PM-8f8ba5e/best_model.pth.tar

EPOCH: 1/1000

Number of output frames: 5

TRAINING (2020-06-07 19:06:01)

–> STEP: 132/166 – GLOBAL_STEP: 300
| > decoder_loss: 0.96331 (1.23226)
| > postnet_loss: 0.92254 (1.31572)
| > stopnet_loss: 0.23428 (0.43323)
| > ga_loss: 0.02817 (0.07714)
| > loss: 1.91402
| > align_error: 0.98439 (0.95954)
| > avg_spec_len: 610.546875
| > avg_text_len: 111.125
| > step_time: 0.84
| > loader_time: 0.02
| > lr: 0.00010
Traceback (most recent call last):
File “train.py”, line 676, in
main(args)
File “train.py”, line 591, in main
global_step, epoch)
File “train.py”, line 170, in train
text_input, text_lengths, mel_input, speaker_ids=speaker_ids)
File “/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py”, line 550, in call
result = self.forward(*input, **kwargs)
File “/opt/conda/lib/python3.7/site-packages/TTS-0.0.2+8f8ba5e-py3.7.egg/TTS/models/tacotron2.py”, line 76, in forward
postnet_outputs = self.postnet(decoder_outputs)
File “/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py”, line 550, in call
result = self.forward(*input, **kwargs)
File “/opt/conda/lib/python3.7/site-packages/TTS-0.0.2+8f8ba5e-py3.7.egg/TTS/layers/tacotron2.py”, line 49, in forward
o = layer(o)
File “/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py”, line 550, in call
result = self.forward(*input, **kwargs)
File “/opt/conda/lib/python3.7/site-packages/TTS-0.0.2+8f8ba5e-py3.7.egg/TTS/layers/tacotron2.py”, line 27, in forward
o = self.convolution1d(x)
File “/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py”, line 550, in call
result = self.forward(*input, **kwargs)
File “/opt/conda/lib/python3.7/site-packages/torch/nn/modules/conv.py”, line 208, in forward
self.padding, self.dilation, self.groups)
RuntimeError: CUDA out of memory. Tried to allocate 250.00 MiB (GPU 0; 15.78 GiB total capacity; 14.32 GiB already allocated; 197.88 MiB free; 14.60 GiB reserved in total by PyTorch)
! Run is kept in …/LJSpeech/ljspeech-June-07-2020_06+46PM-8f8ba5e

log_4_RTX_2080_Ti

Using CUDA: True
Number of GPUs: 4
Git Hash: 8ec568b
Experiment folder: …/LJSpeech/ljspeech-June-07-2020_05+54PM-8ec568b
Setting up Audio Processor…
| > sample_rate:22050
| > num_mels:80
| > min_level_db:-100
| > frame_shift_ms:None
| > frame_length_ms:None
| > ref_level_db:20
| > num_freq:513
| > power:1.5
| > preemphasis:0.0
| > griffin_lim_iters:60
| > signal_norm:True
| > symmetric_norm:True
| > mel_fmin:0
| > mel_fmax:8000.0
| > max_norm:4.0
| > clip_norm:True
| > do_trim_silence:True
| > trim_db:60
| > do_sound_norm:False
| > stats_path:None
| > hop_length:256
| > win_length:1024
| > n_fft:1024
Using model: Tacotron2
| > Num output units : 513

Model has 28921234 parameters

EPOCH: 0/1000

Number of output frames: 7

DataLoader initialization
| > Use phonemes: True
| > phoneme language: it
| > Number of instances : 14820
| > Max length sequence: 1029
| > Min length sequence: 2
| > Avg length sequence: 113.39838056680162
| > Num. instances discarded by max-min (max=153, min=6) seq limits: 4153
| > Batch group size: 0.

TRAINING (2020-06-07 17:54:21)
/opt/conda/lib/python3.7/site-packages/torch/distributed/distributed_c10d.py:102: UserWarning: torch.distributed.reduce_op is deprecated, please use torch.distributed.ReduceOp instead
warnings.warn("torch.distributed.reduce_op is deprecated, please use "
/opt/conda/lib/python3.7/site-packages/torch/distributed/distributed_c10d.py:102: UserWarning: torch.distributed.reduce_op is deprecated, please use torch.distributed.ReduceOp instead
warnings.warn("torch.distributed.reduce_op is deprecated, please use "
/opt/conda/lib/python3.7/site-packages/torch/distributed/distributed_c10d.py:102: UserWarning: torch.distributed.reduce_op is deprecated, please use torch.distributed.ReduceOp instead
warnings.warn("torch.distributed.reduce_op is deprecated, please use "
/opt/conda/lib/python3.7/site-packages/torch/distributed/distributed_c10d.py:102: UserWarning: torch.distributed.reduce_op is deprecated, please use torch.distributed.ReduceOp instead
warnings.warn("torch.distributed.reduce_op is deprecated, please use "

–> STEP: 24/41 – GLOBAL_STEP: 25
| > decoder_loss: 4.34630 (5.94781)
| > postnet_loss: 6.14363 (7.89377)
| > stopnet_loss: 0.70103 (0.75914)
| > ga_loss: 0.03984 (0.10998)
| > loss: 10.52977
| > align_error: 0.98900 (0.96387)
| > avg_spec_len: 438.640625
| > avg_text_len: 77.015625
| > step_time: 6.44
| > loader_time: 33.55
| > lr: 0.00010
[WARNING] fount 1 utterances containing language switches on lines 1
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)
[WARNING] fount 1 utterances containing language switches on lines 2
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)
[WARNING] fount 1 utterances containing language switches on lines 2
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)
[WARNING] fount 1 utterances containing language switches on lines 4
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)
Traceback (most recent call last):
File “train.py”, line 676, in
main(args)
File “train.py”, line 591, in main
global_step, epoch)
File “train.py”, line 170, in train
text_input, text_lengths, mel_input, speaker_ids=speaker_ids)
File “/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py”, line 550, in call
result = self.forward(*input, **kwargs)
File “/opt/conda/lib/python3.7/site-packages/TTS-0.0.2+8ec568b-py3.7.egg/TTS/models/tacotron2.py”, line 75, in forward
encoder_outputs, mel_specs, mask)
File “/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py”, line 550, in call
result = self.forward(*input, **kwargs)
File “/opt/conda/lib/python3.7/site-packages/TTS-0.0.2+8ec568b-py3.7.egg/TTS/layers/tacotron2.py”, line 264, in forward
decoder_output, attention_weights, stop_token = self.decode(memory)
File “/opt/conda/lib/python3.7/site-packages/TTS-0.0.2+8ec568b-py3.7.egg/TTS/layers/tacotron2.py”, line 226, in decode
self.processed_inputs, self.mask)
File “/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py”, line 550, in call
result = self.forward(*input, **kwargs)
File “/opt/conda/lib/python3.7/site-packages/TTS-0.0.2+8ec568b-py3.7.egg/TTS/layers/common_layers.py”, line 335, in forward
query, processed_inputs)
File “/opt/conda/lib/python3.7/site-packages/TTS-0.0.2+8ec568b-py3.7.egg/TTS/layers/common_layers.py”, line 275, in get_location_attention
torch.tanh(processed_query + processed_attention_weights +
RuntimeError: CUDA out of memory. Tried to allocate 18.00 MiB (GPU 2; 10.76 GiB total capacity; 9.86 GiB already allocated; 17.56 MiB free; 9.96 GiB reserved in total by PyTorch)

6_RTX_2080_Ti

Using CUDA: True
Number of GPUs: 6
Git Hash: 8f8ba5e
Experiment folder: …/LJSpeech/ljspeech-June-07-2020_09+10PM-8f8ba5e
Setting up Audio Processor…
| > sample_rate:22050
| > num_mels:80
| > min_level_db:-100
| > frame_shift_ms:None
| > frame_length_ms:None
| > ref_level_db:20
| > num_freq:513
| > power:1.5
| > preemphasis:0.0
| > griffin_lim_iters:60
| > signal_norm:True
| > symmetric_norm:True
| > mel_fmin:0
| > mel_fmax:8000.0
| > max_norm:4.0
| > clip_norm:True
| > do_trim_silence:True
| > trim_db:60
| > do_sound_norm:False
| > stats_path:None
| > hop_length:256
| > win_length:1024
| > n_fft:1024
Using model: Tacotron2
| > Num output units : 513

Model has 28921234 parameters

EPOCH: 0/1000

Number of output frames: 7

DataLoader initialization
| > Use phonemes: True
| > phoneme language: it
| > Number of instances : 14820
| > Max length sequence: 1029
| > Min length sequence: 2
| > Avg length sequence: 113.39838056680162
| > Num. instances discarded by max-min (max=153, min=6) seq limits: 4153
| > Batch group size: 0.

TRAINING (2020-06-07 21:10:31)
/opt/conda/lib/python3.7/site-packages/torch/distributed/distributed_c10d.py:102: UserWarning: torch.distributed.reduce_op is deprecated, please use torch.distributed.ReduceOp instead
warnings.warn("torch.distributed.reduce_op is deprecated, please use "
/opt/conda/lib/python3.7/site-packages/torch/distributed/distributed_c10d.py:102: UserWarning: torch.distributed.reduce_op is deprecated, please use torch.distributed.ReduceOp instead
warnings.warn("torch.distributed.reduce_op is deprecated, please use "
/opt/conda/lib/python3.7/site-packages/torch/distributed/distributed_c10d.py:102: UserWarning: torch.distributed.reduce_op is deprecated, please use torch.distributed.ReduceOp instead
warnings.warn("torch.distributed.reduce_op is deprecated, please use "
/opt/conda/lib/python3.7/site-packages/torch/distributed/distributed_c10d.py:102: UserWarning: torch.distributed.reduce_op is deprecated, please use torch.distributed.ReduceOp instead
warnings.warn("torch.distributed.reduce_op is deprecated, please use "
/opt/conda/lib/python3.7/site-packages/torch/distributed/distributed_c10d.py:102: UserWarning: torch.distributed.reduce_op is deprecated, please use torch.distributed.ReduceOp instead
warnings.warn("torch.distributed.reduce_op is deprecated, please use "
/opt/conda/lib/python3.7/site-packages/torch/distributed/distributed_c10d.py:102: UserWarning: torch.distributed.reduce_op is deprecated, please use torch.distributed.ReduceOp instead
warnings.warn("torch.distributed.reduce_op is deprecated, please use "
[WARNING] fount 1 utterances containing language switches on lines 2
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)
[WARNING] fount 1 utterances containing language switches on lines 1
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)
[WARNING] fount 1 utterances containing language switches on lines 4
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)
[WARNING] fount 1 utterances containing language switches on lines 2
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)
Traceback (most recent call last):
File “train.py”, line 676, in
main(args)
File “train.py”, line 591, in main
global_step, epoch)
File “train.py”, line 170, in train
text_input, text_lengths, mel_input, speaker_ids=speaker_ids)
File “/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py”, line 550, in call
result = self.forward(*input, **kwargs)
File “/opt/conda/lib/python3.7/site-packages/TTS-0.0.2+8f8ba5e-py3.7.egg/TTS/models/tacotron2.py”, line 75, in forward
encoder_outputs, mel_specs, mask)
File “/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py”, line 550, in call
result = self.forward(*input, **kwargs)
File “/opt/conda/lib/python3.7/site-packages/TTS-0.0.2+8f8ba5e-py3.7.egg/TTS/layers/tacotron2.py”, line 264, in forward
decoder_output, attention_weights, stop_token = self.decode(memory)
File “/opt/conda/lib/python3.7/site-packages/TTS-0.0.2+8f8ba5e-py3.7.egg/TTS/layers/tacotron2.py”, line 226, in decode
self.processed_inputs, self.mask)
File “/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py”, line 550, in call
result = self.forward(*input, **kwargs)
File “/opt/conda/lib/python3.7/site-packages/TTS-0.0.2+8f8ba5e-py3.7.egg/TTS/layers/common_layers.py”, line 335, in forward
query, processed_inputs)
File “/opt/conda/lib/python3.7/site-packages/TTS-0.0.2+8f8ba5e-py3.7.egg/TTS/layers/common_layers.py”, line 276, in get_location_attention
processed_inputs))
RuntimeError: CUDA out of memory. Tried to allocate 18.00 MiB (GPU 0; 10.76 GiB total capacity; 9.87 GiB already allocated; 10.25 MiB free; 9.97 GiB reserved in total by PyTorch)
! Run is removed from …/LJSpeech/ljspeech-June-07-2020_09+10PM-8f8ba5e
[WARNING] fount 1 utterances containing language switches on lines 2
[WARNING] extra phones may appear in the “it” phoneset
[WARNING] language switch flags have been kept (applying “keep-flags” policy)

erogol · June 8, 2020, 9:23am

trim long sampled shorter or use smaller batch size.

Sadam1195 · June 8, 2020, 9:35am

Thanks for quick response. I figured using trimmed sentences with shorter duration might work, I’ll try that and post my response back. Although I did try using smaller batch size but it didn’t work.