eden-emu/opus
15
0
Fork 0
mirror of https://github.com/xiph/opus.git synced 2025-05-18 01:18:30 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 4

moving hyper-parameters to new_lpcnet_model() arguments

Browse source 
more cleaning up too
This commit is contained in:
Jean-Marc Valin 2018-10-24 14:57:34 -04:00
parent 5ff58fa117
commit db7569c3da
3 changed files with 11 additions and 15 deletions
Download patch file Download diff file Expand all files Collapse all files

9
dnn/lpcnet.py
View file

@ -11,12 +11,9 @@ import numpy as np
import h5py import h5py
import sys import sys
rnn_units1=384
rnn_units2=16
pcm_bits = 8 pcm_bits = 8
embed_size = 128 embed_size = 128
pcm_levels = 2**pcm_bits pcm_levels = 2**pcm_bits
nb_used_features = 38
class Sparsify(Callback): class Sparsify(Callback):
def __init__(self, t_start, t_end, interval, density): def __init__(self, t_start, t_end, interval, density):
@ -88,7 +85,7 @@ class PCMInit(Initializer):
'seed': self.seed 'seed': self.seed
} }
def new_wavernn_model(): def new_lpcnet_model(rnn_units1=384, rnn_units2=16, nb_used_features = 38):
pcm = Input(shape=(None, 2)) pcm = Input(shape=(None, 2))
exc = Input(shape=(None, 1)) exc = Input(shape=(None, 1))
feat = Input(shape=(None, nb_used_features)) feat = Input(shape=(None, nb_used_features))
@ -127,6 +124,10 @@ def new_wavernn_model():
ulaw_prob = md(gru_out2) ulaw_prob = md(gru_out2)
model = Model([pcm, exc, feat, pitch], ulaw_prob) model = Model([pcm, exc, feat, pitch], ulaw_prob)
model.rnn_units1 = rnn_units1
model.rnn_units2 = rnn_units2
model.nb_used_features = nb_used_features
encoder = Model([feat, pitch], cfeat) encoder = Model([feat, pitch], cfeat)
dec_rnn_in = Concatenate()([cpcm, cexc, dec_feat]) dec_rnn_in = Concatenate()([cpcm, cexc, dec_feat])

12
dnn/test_lpcnet.py
View file

@ -15,11 +15,7 @@ config = tf.ConfigProto()
config.gpu_options.per_process_gpu_memory_fraction = 0.2 config.gpu_options.per_process_gpu_memory_fraction = 0.2
set_session(tf.Session(config=config)) set_session(tf.Session(config=config))
nb_epochs = 40 model, enc, dec = lpcnet.new_lpcnet_model()
batch_size = 64
#model = wavenet.new_wavenet_model(fftnet=True)
model, enc, dec = lpcnet.new_wavernn_model()
model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['sparse_categorical_accuracy']) model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['sparse_categorical_accuracy'])
#model.summary() #model.summary()
@ -27,7 +23,7 @@ model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=
feature_file = sys.argv[1] feature_file = sys.argv[1]
frame_size = 160 frame_size = 160
nb_features = 55 nb_features = 55
nb_used_features = lpcnet.nb_used_features nb_used_features = model.nb_used_features
features = np.fromfile(feature_file, dtype='float32') features = np.fromfile(feature_file, dtype='float32')
features = np.resize(features, (-1, nb_features)) features = np.resize(features, (-1, nb_features))
@ -48,8 +44,8 @@ order = 16
pcm = np.zeros((nb_frames*pcm_chunk_size, )) pcm = np.zeros((nb_frames*pcm_chunk_size, ))
fexc = np.zeros((1, 1, 2), dtype='float32') fexc = np.zeros((1, 1, 2), dtype='float32')
iexc = np.zeros((1, 1, 1), dtype='int16') iexc = np.zeros((1, 1, 1), dtype='int16')
state1 = np.zeros((1, lpcnet.rnn_units1), dtype='float32') state1 = np.zeros((1, model.rnn_units1), dtype='float32')
state2 = np.zeros((1, lpcnet.rnn_units2), dtype='float32') state2 = np.zeros((1, model.rnn_units2), dtype='float32')
mem = 0 mem = 0
coef = 0.85 coef = 0.85

5
dnn/train_lpcnet.py
View file

@ -28,8 +28,7 @@ nb_epochs = 40
# Try reducing batch_size if you run out of memory on your GPU # Try reducing batch_size if you run out of memory on your GPU
batch_size = 64 batch_size = 64
# Note we are creating a LPCNet model model, _, _ = lpcnet.new_lpcnet_model()
model, _, _ = lpcnet.new_wavernn_model()
model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['sparse_categorical_accuracy']) model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['sparse_categorical_accuracy'])
model.summary() model.summary()
@ -40,7 +39,7 @@ pred_file = sys.argv[3] # LPC predictor samples. Not used at present, see bel
pcm_file = sys.argv[4] # 16 bit unsigned short PCM samples pcm_file = sys.argv[4] # 16 bit unsigned short PCM samples
frame_size = 160 frame_size = 160
nb_features = 55 nb_features = 55
nb_used_features = lpcnet.nb_used_features nb_used_features = model.nb_used_features
feature_chunk_size = 15 feature_chunk_size = 15
pcm_chunk_size = frame_size*feature_chunk_size pcm_chunk_size = frame_size*feature_chunk_size