hub / github.com/NVIDIA/waveglow / train

Function train

train.py:62–152 · view source on GitHub ↗

(num_gpus, rank, group_name, output_directory, epochs, learning_rate,
          sigma, iters_per_checkpoint, batch_size, seed, fp16_run,
          checkpoint_path, with_tensorboard)

Source from the content-addressed store, hash-verified

source not stored for this graph (policy: none)

Callers 1

train.pyFile · 0.85

Calls 8

init_distributedFunction · 0.90

WaveGlowLossClass · 0.90

WaveGlowClass · 0.90

apply_gradient_allreduceFunction · 0.90

Mel2SampClass · 0.90

reduce_tensorFunction · 0.90

load_checkpointFunction · 0.85

save_checkpointFunction · 0.85

Tested by

no test coverage detected