200413d subset OK but all_8k rough

src/700c_train.sh

@@ -17,7 +17,7 @@ test1=all_speech_subset_8k
 test2=all_8k
 test3="birch canadian glue oak separately wanted wia peter"
 datestamp=$1
-epochs=5
+epochs=10
 log=${1}.txt
 train=${datestamp}_train
 
@@ -66,19 +66,20 @@ experiment() {
 
 (
     rm -f $log
-
+    
     # assemble some training speech
     sox -r 8000 -c 1 ~/Downloads/${train1}.sw \
     -r 8000 -c 1 ~/Downloads/${train2}.sw \
     -t sw -r 8000 -c 1 ${train}.sw
     #cp ~/Downloads/${train1}.sw ${train}.sw  
-
+    : '
     experiment "" "none"           # no prediction
 
     synth_40ms ${test1}    
     synth_40ms ${test2}
     
     experiment "--first"  "first" # first order predictor
+    '
     experiment "--lpc 10" "lpc"   # standard LPC (albiet 10th order)
 
 ) |& tee $log

src/train_lpcnet.py

@@ -63,6 +63,7 @@ parser.add_argument('prefix', help='.h5 file prefix to easily identify each expe
 parser.add_argument('--frame_size', type=int, default=160, help='frames size in samples')
 parser.add_argument('--epochs', type=int, default=20, help='Number of training epochs')
 parser.add_argument('--no_pitch_embedding', action='store_true', help='disable pitch embedding')
+parser.add_argument('--load_h5', help='disable pitch embedding')
 args = parser.parse_args()
 
 nb_epochs = args.epochs
@@ -72,6 +73,10 @@ model, _, _ = lpcnet.new_lpcnet_model(frame_size=args.frame_size, training=True)
 model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['sparse_categorical_accuracy'])
 model.summary()
 
+if args.load_h5:
+    print("loading: %s" % (args.load_h5))
+    model.load_weights(args.load_h5)
+
 feature_file = args.feature_file
 pcm_file = args.packed_ulaw_file           
 prefix = args.prefix              
@@ -157,6 +162,5 @@ del in_exc
 checkpoint = ModelCheckpoint(prefix + '_{epoch:d}.h5')
 
 # use this to reload a partially trained model
-#model.load_weights('lpcnet_190203_07.h5')
 model.compile(optimizer=Adam(0.001, amsgrad=True, decay=5e-5), loss='sparse_categorical_crossentropy')
 model.fit([in_data, features, periods], out_exc, batch_size=batch_size, epochs=nb_epochs, callbacks=[checkpoint, lpcnet.Sparsify(2000, 40000, 400, (0.05, 0.05, 0.2))])