Upload SSL Layer MOS checkpoints (part 6)

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run2/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d60cd3794de1d66baa13614d24fe087acd81193f9575d6dee0eafb8cc86fee95
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run2/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c148ec4598b2a0ca2c6a483bd0ed2850d30262cbfdc00d8802e714bbdbbe07a7
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run2/train.log

@@ -0,0 +1,99 @@
+INFO:root:Num train speech clips: 4974
+INFO:root:Num val speech clips: 1066
+INFO:root:Num test speech clips: 1066
+INFO:root:Device=cuda
+INFO:root:Average loss=0.48212505896122027
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3767 | LCC = 0.8105 | SRCC = 0.8180 ]
+INFO:root:Average loss=0.24159042499004266
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3070 | LCC = 0.8531 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.21392513028322122
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2711 | LCC = 0.8600 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.19145774669372118
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2558 | LCC = 0.8664 | SRCC = 0.8677 ]
+INFO:root:Average loss=0.17701337257256874
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2499 | LCC = 0.8706 | SRCC = 0.8708 ]
+INFO:root:Average loss=0.1668390225714598
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2997 | LCC = 0.8722 | SRCC = 0.8724 ]
+INFO:root:Average loss=0.1563700186327482
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2658 | LCC = 0.8744 | SRCC = 0.8728 ]
+INFO:root:Average loss=0.14279718085741386
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2362 | LCC = 0.8737 | SRCC = 0.8735 ]
+INFO:root:Average loss=0.1286015476171787
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2080 | LCC = 0.8730 | SRCC = 0.8730 ]
+INFO:root:Average loss=0.1253895195057759
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2415 | LCC = 0.8714 | SRCC = 0.8709 ]
+INFO:root:Average loss=0.119387052857723
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3019 | LCC = 0.8696 | SRCC = 0.8730 ]
+INFO:root:Average loss=0.10998855263758929
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2433 | LCC = 0.8676 | SRCC = 0.8677 ]
+INFO:root:Average loss=0.10329175329743287
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2288 | LCC = 0.8697 | SRCC = 0.8711 ]
+INFO:root:Average loss=0.0967078777268911
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2480 | LCC = 0.8678 | SRCC = 0.8681 ]
+INFO:root:Average loss=0.0947616626150333
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3041 | LCC = 0.8686 | SRCC = 0.8702 ]
+INFO:root:Average loss=0.08734149084641384
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2650 | LCC = 0.8690 | SRCC = 0.8692 ]
+INFO:root:Average loss=0.08210476965476306
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2705 | LCC = 0.8658 | SRCC = 0.8680 ]
+INFO:root:Average loss=0.0858790724991988
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2605 | LCC = 0.8669 | SRCC = 0.8683 ]
+INFO:root:Average loss=0.08271859643551019
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2400 | LCC = 0.8669 | SRCC = 0.8677 ]
+INFO:root:Average loss=0.07146724886619128
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2449 | LCC = 0.8641 | SRCC = 0.8661 ]
+INFO:root:Average loss=0.06925657678108949
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2343 | LCC = 0.8644 | SRCC = 0.8657 ]
+INFO:root:Average loss=0.06704354940507656
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2738 | LCC = 0.8642 | SRCC = 0.8652 ]
+INFO:root:Average loss=0.07050568801470292
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2668 | LCC = 0.8640 | SRCC = 0.8649 ]
+INFO:root:Average loss=0.06089034466407238
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2396 | LCC = 0.8632 | SRCC = 0.8643 ]
+INFO:root:Average loss=0.05812540471267242
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2460 | LCC = 0.8622 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.05923682924073476
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2718 | LCC = 0.8634 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.056070149183655396
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2424 | LCC = 0.8621 | SRCC = 0.8632 ]
+INFO:root:Average loss=0.052097282253014736
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2522 | LCC = 0.8628 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.051188175638134666
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2847 | LCC = 0.8603 | SRCC = 0.8611 ]
+INFO:root:Average loss=0.05074002846884422
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2526 | LCC = 0.8610 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.048052534532661624
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2647 | LCC = 0.8614 | SRCC = 0.8628 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2438 | LCC = 0.8729 | SRCC = 0.8723 ]

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run3/config.gin

@@ -0,0 +1,31 @@
+# Training config on BVCC dataset.
+
+# ------------------------------
+# Architecture
+# ------------------------------
+ProjectionHead:
+    conv_channels = (32, 32)
+    dense_neurons = (64, 32, 1)
+    use_poolings = (True, True)
+    kernel_size = 5
+    pool_size = 5
+    
+# ------------------------------
+# Dataset
+# ------------------------------
+Bvcc:
+    debug = False
+
+get_dataloader:
+    num_workers = 12
+    shuffle = True
+
+# ------------------------------
+# Training loop
+# ------------------------------
+TrainingLoop:
+    num_epochs = 30
+    learning_rate = 1e-4
+    dataset_cls = @Bvcc
+    weight_decay = 0.0
+    batch_size_train = 64

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run3/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3c9621772466bf7b22af2c71b392b66791293e07e591edf5c88a136fa4acff0
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run3/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c2920e2ab5cd64fb24703afe28923daef978fdd28d3f36fca9bf2eefb83c8f4
+size 2749638

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run3/train.log

@@ -0,0 +1,99 @@
+INFO:root:Num train speech clips: 4974
+INFO:root:Num val speech clips: 1066
+INFO:root:Num test speech clips: 1066
+INFO:root:Device=cuda
+INFO:root:Average loss=0.403517763202007
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3261 | LCC = 0.8120 | SRCC = 0.8157 ]
+INFO:root:Average loss=0.24763199935356775
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2886 | LCC = 0.8439 | SRCC = 0.8436 ]
+INFO:root:Average loss=0.21508449908250418
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2625 | LCC = 0.8551 | SRCC = 0.8547 ]
+INFO:root:Average loss=0.20363809244754988
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2662 | LCC = 0.8604 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.1909643248296701
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2431 | LCC = 0.8619 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.16637271384780222
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2367 | LCC = 0.8639 | SRCC = 0.8628 ]
+INFO:root:Average loss=0.16145546495532379
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2372 | LCC = 0.8634 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.14887363778857085
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2355 | LCC = 0.8636 | SRCC = 0.8640 ]
+INFO:root:Average loss=0.1447817728114434
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2328 | LCC = 0.8648 | SRCC = 0.8654 ]
+INFO:root:Average loss=0.1334422770410012
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2355 | LCC = 0.8622 | SRCC = 0.8635 ]
+INFO:root:Average loss=0.12572158624728522
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2291 | LCC = 0.8623 | SRCC = 0.8632 ]
+INFO:root:Average loss=0.12735722194879484
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2256 | LCC = 0.8630 | SRCC = 0.8648 ]
+INFO:root:Average loss=0.1120024187824665
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2372 | LCC = 0.8591 | SRCC = 0.8614 ]
+INFO:root:Average loss=0.10663311761350204
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2204 | LCC = 0.8637 | SRCC = 0.8666 ]
+INFO:root:Average loss=0.10106851475743148
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2390 | LCC = 0.8581 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.0962457336867467
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2224 | LCC = 0.8619 | SRCC = 0.8650 ]
+INFO:root:Average loss=0.09278945452891864
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2267 | LCC = 0.8607 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.08597430725319263
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2215 | LCC = 0.8616 | SRCC = 0.8640 ]
+INFO:root:Average loss=0.08565764367962495
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2293 | LCC = 0.8603 | SRCC = 0.8628 ]
+INFO:root:Average loss=0.07930731133390696
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2194 | LCC = 0.8614 | SRCC = 0.8628 ]
+INFO:root:Average loss=0.07558726586210422
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2230 | LCC = 0.8621 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.07274589644601712
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2332 | LCC = 0.8597 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.07222348241469799
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2285 | LCC = 0.8617 | SRCC = 0.8630 ]
+INFO:root:Average loss=0.06443445074061553
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2423 | LCC = 0.8543 | SRCC = 0.8568 ]
+INFO:root:Average loss=0.06261538699842416
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2381 | LCC = 0.8573 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.061561480355568424
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2324 | LCC = 0.8597 | SRCC = 0.8617 ]
+INFO:root:Average loss=0.05873171679484539
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2217 | LCC = 0.8598 | SRCC = 0.8619 ]
+INFO:root:Average loss=0.059468637817563154
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2265 | LCC = 0.8576 | SRCC = 0.8604 ]
+INFO:root:Average loss=0.05554105355762518
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2261 | LCC = 0.8598 | SRCC = 0.8625 ]
+INFO:root:Average loss=0.053816242525592826
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2218 | LCC = 0.8609 | SRCC = 0.8635 ]
+INFO:root:Average loss=0.05056486806521813
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2319 | LCC = 0.8578 | SRCC = 0.8594 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2384 | LCC = 0.8679 | SRCC = 0.8661 ]

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run4/config.gin

@@ -0,0 +1,31 @@
+# Training config on BVCC dataset.
+
+# ------------------------------
+# Architecture
+# ------------------------------
+ProjectionHead:
+    conv_channels = (32, 32)
+    dense_neurons = (64, 32, 1)
+    use_poolings = (True, True)
+    kernel_size = 5
+    pool_size = 5
+    
+# ------------------------------
+# Dataset
+# ------------------------------
+Bvcc:
+    debug = False
+
+get_dataloader:
+    num_workers = 12
+    shuffle = True
+
+# ------------------------------
+# Training loop
+# ------------------------------
+TrainingLoop:
+    num_epochs = 30
+    learning_rate = 1e-4
+    dataset_cls = @Bvcc
+    weight_decay = 0.0
+    batch_size_train = 64

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run4/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86ccb06f8c74689c38a6844be54c98d2e987accabae067549783af6952b76ba1
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run4/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da524cb78c9911a150206922505ae3c790bbc870ed7c69552fe3963a52460c04
+size 2749638

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run4/train.log

@@ -0,0 +1,99 @@
+INFO:root:Num train speech clips: 4974
+INFO:root:Num val speech clips: 1066
+INFO:root:Num test speech clips: 1066
+INFO:root:Device=cuda
+INFO:root:Average loss=0.4551508388458154
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3547 | LCC = 0.8126 | SRCC = 0.8195 ]
+INFO:root:Average loss=0.24563788699033934
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2633 | LCC = 0.8477 | SRCC = 0.8490 ]
+INFO:root:Average loss=0.22253791624918962
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2699 | LCC = 0.8594 | SRCC = 0.8589 ]
+INFO:root:Average loss=0.20261611082614997
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2980 | LCC = 0.8631 | SRCC = 0.8631 ]
+INFO:root:Average loss=0.18679375287431937
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2468 | LCC = 0.8674 | SRCC = 0.8662 ]
+INFO:root:Average loss=0.1766640214392772
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2959 | LCC = 0.8662 | SRCC = 0.8662 ]
+INFO:root:Average loss=0.16893091501715857
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2464 | LCC = 0.8692 | SRCC = 0.8684 ]
+INFO:root:Average loss=0.15587023731607658
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2492 | LCC = 0.8702 | SRCC = 0.8696 ]
+INFO:root:Average loss=0.14170436952740717
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2697 | LCC = 0.8688 | SRCC = 0.8693 ]
+INFO:root:Average loss=0.14151361517799207
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2423 | LCC = 0.8704 | SRCC = 0.8697 ]
+INFO:root:Average loss=0.1240942660623636
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2586 | LCC = 0.8640 | SRCC = 0.8648 ]
+INFO:root:Average loss=0.12205112629976028
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2629 | LCC = 0.8679 | SRCC = 0.8679 ]
+INFO:root:Average loss=0.11233795577516922
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2502 | LCC = 0.8637 | SRCC = 0.8640 ]
+INFO:root:Average loss=0.10628868582156989
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2220 | LCC = 0.8651 | SRCC = 0.8657 ]
+INFO:root:Average loss=0.09721051949339035
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2459 | LCC = 0.8624 | SRCC = 0.8628 ]
+INFO:root:Average loss=0.09455199577869514
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2625 | LCC = 0.8611 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.08903195842718467
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2253 | LCC = 0.8645 | SRCC = 0.8648 ]
+INFO:root:Average loss=0.08365719287823407
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2281 | LCC = 0.8665 | SRCC = 0.8670 ]
+INFO:root:Average loss=0.08212937830159298
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2630 | LCC = 0.8632 | SRCC = 0.8663 ]
+INFO:root:Average loss=0.08240758040203498
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2274 | LCC = 0.8636 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.07459291448004735
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2406 | LCC = 0.8621 | SRCC = 0.8625 ]
+INFO:root:Average loss=0.07192672082246879
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2720 | LCC = 0.8624 | SRCC = 0.8635 ]
+INFO:root:Average loss=0.07326538469164799
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2480 | LCC = 0.8634 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.062272916141992964
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2515 | LCC = 0.8616 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.06356821720225689
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2588 | LCC = 0.8613 | SRCC = 0.8628 ]
+INFO:root:Average loss=0.06053972760072121
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2596 | LCC = 0.8613 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.05774620189689673
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2594 | LCC = 0.8618 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.06022118270779268
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2782 | LCC = 0.8547 | SRCC = 0.8564 ]
+INFO:root:Average loss=0.055904310244398236
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2515 | LCC = 0.8604 | SRCC = 0.8610 ]
+INFO:root:Average loss=0.052592406049370766
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2473 | LCC = 0.8630 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.051705676632431835
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2558 | LCC = 0.8633 | SRCC = 0.8650 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2411 | LCC = 0.8709 | SRCC = 0.8711 ]

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run0/config.gin

@@ -0,0 +1,31 @@
+# Training config on BVCC dataset.
+
+# ------------------------------
+# Architecture
+# ------------------------------
+ProjectionHead:
+    conv_channels = (32, 32)
+    dense_neurons = (64, 32, 1)
+    use_poolings = (True, True)
+    kernel_size = 5
+    pool_size = 5
+    
+# ------------------------------
+# Dataset
+# ------------------------------
+Bvcc:
+    debug = False
+
+get_dataloader:
+    num_workers = 12
+    shuffle = True
+
+# ------------------------------
+# Training loop
+# ------------------------------
+TrainingLoop:
+    num_epochs = 30
+    learning_rate = 1e-4
+    dataset_cls = @Bvcc
+    weight_decay = 0.0
+    batch_size_train = 64

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run0/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ebcc46cd859d8ac4d69cea6d604adbabe0cbebb9714c8c4d5d0cce6fd05a25c5
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run0/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:607e786e0f20d96097642cab6ea07da63324a16113243dbb66953965875a7c64
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run0/train.log

@@ -0,0 +1,99 @@
+INFO:root:Num train speech clips: 4974
+INFO:root:Num val speech clips: 1066
+INFO:root:Num test speech clips: 1066
+INFO:root:Device=cuda
+INFO:root:Average loss=0.3572192581800314
+INFO:root:
+[Valid][0][UTT][ MSE = 0.2990 | LCC = 0.8250 | SRCC = 0.8294 ]
+INFO:root:Average loss=0.2406188360391519
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2557 | LCC = 0.8495 | SRCC = 0.8508 ]
+INFO:root:Average loss=0.21774951502298698
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2520 | LCC = 0.8550 | SRCC = 0.8555 ]
+INFO:root:Average loss=0.1942141281488614
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2529 | LCC = 0.8609 | SRCC = 0.8614 ]
+INFO:root:Average loss=0.18217804454840147
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2318 | LCC = 0.8649 | SRCC = 0.8642 ]
+INFO:root:Average loss=0.16384037488546127
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2206 | LCC = 0.8636 | SRCC = 0.8641 ]
+INFO:root:Average loss=0.15677847331150985
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2326 | LCC = 0.8631 | SRCC = 0.8614 ]
+INFO:root:Average loss=0.14156064238303748
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2275 | LCC = 0.8631 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.12919007184413764
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2446 | LCC = 0.8625 | SRCC = 0.8640 ]
+INFO:root:Average loss=0.12635595399217728
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2200 | LCC = 0.8653 | SRCC = 0.8662 ]
+INFO:root:Average loss=0.11152979253958432
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2488 | LCC = 0.8591 | SRCC = 0.8591 ]
+INFO:root:Average loss=0.10957995362770863
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2467 | LCC = 0.8597 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.10441889871771519
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2639 | LCC = 0.8570 | SRCC = 0.8600 ]
+INFO:root:Average loss=0.09284338278648181
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2341 | LCC = 0.8575 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.09441712059271641
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2377 | LCC = 0.8574 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.08448655019776943
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2315 | LCC = 0.8576 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.08097663440574439
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2352 | LCC = 0.8553 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.07481831455459961
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2270 | LCC = 0.8577 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.07260157569096638
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2393 | LCC = 0.8541 | SRCC = 0.8559 ]
+INFO:root:Average loss=0.06735953172812095
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2560 | LCC = 0.8505 | SRCC = 0.8516 ]
+INFO:root:Average loss=0.06430891509621571
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2430 | LCC = 0.8521 | SRCC = 0.8521 ]
+INFO:root:Average loss=0.06036842332627529
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2381 | LCC = 0.8530 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.057029750174245775
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2371 | LCC = 0.8524 | SRCC = 0.8524 ]
+INFO:root:Average loss=0.054139966336198345
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2436 | LCC = 0.8494 | SRCC = 0.8511 ]
+INFO:root:Average loss=0.058067966085404925
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2339 | LCC = 0.8524 | SRCC = 0.8533 ]
+INFO:root:Average loss=0.05401303437657845
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2404 | LCC = 0.8510 | SRCC = 0.8514 ]
+INFO:root:Average loss=0.05202999477012035
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2356 | LCC = 0.8527 | SRCC = 0.8538 ]
+INFO:root:Average loss=0.050806347806102194
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2448 | LCC = 0.8505 | SRCC = 0.8508 ]
+INFO:root:Average loss=0.05302146397148951
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2341 | LCC = 0.8533 | SRCC = 0.8541 ]
+INFO:root:Average loss=0.04574633611795994
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2468 | LCC = 0.8508 | SRCC = 0.8516 ]
+INFO:root:Average loss=0.046199703469681434
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2353 | LCC = 0.8528 | SRCC = 0.8535 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2419 | LCC = 0.8587 | SRCC = 0.8598 ]

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run1/config.gin

@@ -0,0 +1,31 @@
+# Training config on BVCC dataset.
+
+# ------------------------------
+# Architecture
+# ------------------------------
+ProjectionHead:
+    conv_channels = (32, 32)
+    dense_neurons = (64, 32, 1)
+    use_poolings = (True, True)
+    kernel_size = 5
+    pool_size = 5
+    
+# ------------------------------
+# Dataset
+# ------------------------------
+Bvcc:
+    debug = False
+
+get_dataloader:
+    num_workers = 12
+    shuffle = True
+
+# ------------------------------
+# Training loop
+# ------------------------------
+TrainingLoop:
+    num_epochs = 30
+    learning_rate = 1e-4
+    dataset_cls = @Bvcc
+    weight_decay = 0.0
+    batch_size_train = 64

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run1/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e00ef37f9cb1ad312ca4939acfc88b337d981cc2b99eec9bec092805be5ebf96
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run1/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da7c506c355900a70621f47dca5a770f72fb8c665061a28c39d1f2eca0048187
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run1/train.log

@@ -0,0 +1,99 @@
+INFO:root:Num train speech clips: 4974
+INFO:root:Num val speech clips: 1066
+INFO:root:Num test speech clips: 1066
+INFO:root:Device=cuda
+INFO:root:Average loss=0.40165887410059953
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3370 | LCC = 0.8110 | SRCC = 0.8176 ]
+INFO:root:Average loss=0.2551034534206757
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2864 | LCC = 0.8380 | SRCC = 0.8388 ]
+INFO:root:Average loss=0.2264320825537046
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2783 | LCC = 0.8516 | SRCC = 0.8510 ]
+INFO:root:Average loss=0.2056539114087056
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2616 | LCC = 0.8583 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.1841818304398121
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2413 | LCC = 0.8612 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.17127456850348374
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2516 | LCC = 0.8613 | SRCC = 0.8588 ]
+INFO:root:Average loss=0.16312490986325803
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2361 | LCC = 0.8636 | SRCC = 0.8625 ]
+INFO:root:Average loss=0.14493517319743449
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2413 | LCC = 0.8646 | SRCC = 0.8640 ]
+INFO:root:Average loss=0.13562019255298835
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2309 | LCC = 0.8662 | SRCC = 0.8655 ]
+INFO:root:Average loss=0.1305906085822827
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2282 | LCC = 0.8639 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.1189008272993259
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2379 | LCC = 0.8627 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.11392348364759715
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2512 | LCC = 0.8626 | SRCC = 0.8625 ]
+INFO:root:Average loss=0.10430541462623157
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2261 | LCC = 0.8638 | SRCC = 0.8648 ]
+INFO:root:Average loss=0.10264057455918728
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2674 | LCC = 0.8569 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.09011854971639621
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2496 | LCC = 0.8565 | SRCC = 0.8568 ]
+INFO:root:Average loss=0.09496126629603215
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2301 | LCC = 0.8611 | SRCC = 0.8604 ]
+INFO:root:Average loss=0.08675182386277577
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2584 | LCC = 0.8585 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.0793410558731128
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2475 | LCC = 0.8587 | SRCC = 0.8590 ]
+INFO:root:Average loss=0.07912692351218982
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3170 | LCC = 0.8556 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.07360794576696861
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2426 | LCC = 0.8587 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.06506750078346485
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2514 | LCC = 0.8573 | SRCC = 0.8566 ]
+INFO:root:Average loss=0.0636703636115178
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2360 | LCC = 0.8598 | SRCC = 0.8604 ]
+INFO:root:Average loss=0.059694678689806886
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2521 | LCC = 0.8571 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.0643480230982487
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2336 | LCC = 0.8582 | SRCC = 0.8588 ]
+INFO:root:Average loss=0.058846390447937526
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2525 | LCC = 0.8565 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.055393779363769755
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2305 | LCC = 0.8596 | SRCC = 0.8599 ]
+INFO:root:Average loss=0.05909108141293892
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2454 | LCC = 0.8560 | SRCC = 0.8564 ]
+INFO:root:Average loss=0.04971969309143531
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2265 | LCC = 0.8580 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.05274124906804317
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2548 | LCC = 0.8561 | SRCC = 0.8575 ]
+INFO:root:Average loss=0.047600057907402515
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2492 | LCC = 0.8569 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.04698935822130014
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2608 | LCC = 0.8570 | SRCC = 0.8583 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2366 | LCC = 0.8565 | SRCC = 0.8588 ]

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run2/config.gin

@@ -0,0 +1,31 @@
+# Training config on BVCC dataset.
+
+# ------------------------------
+# Architecture
+# ------------------------------
+ProjectionHead:
+    conv_channels = (32, 32)
+    dense_neurons = (64, 32, 1)
+    use_poolings = (True, True)
+    kernel_size = 5
+    pool_size = 5
+    
+# ------------------------------
+# Dataset
+# ------------------------------
+Bvcc:
+    debug = False
+
+get_dataloader:
+    num_workers = 12
+    shuffle = True
+
+# ------------------------------
+# Training loop
+# ------------------------------
+TrainingLoop:
+    num_epochs = 30
+    learning_rate = 1e-4
+    dataset_cls = @Bvcc
+    weight_decay = 0.0
+    batch_size_train = 64

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run2/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:deb21e92806edde8f1636efcf72a3ade7392f0b8d4a615147811b20f6452e72f
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run2/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb13be8e584ae3c957cfde558aab4bbb9a6cf2ae18319fa3bb60d14dbeccc10c
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run2/train.log

@@ -0,0 +1,99 @@
+INFO:root:Num train speech clips: 4974
+INFO:root:Num val speech clips: 1066
+INFO:root:Num test speech clips: 1066
+INFO:root:Device=cuda
+INFO:root:Average loss=0.35169577025450194
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3018 | LCC = 0.8219 | SRCC = 0.8307 ]
+INFO:root:Average loss=0.24696448483528236
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2837 | LCC = 0.8493 | SRCC = 0.8482 ]
+INFO:root:Average loss=0.2105240334685032
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2623 | LCC = 0.8594 | SRCC = 0.8617 ]
+INFO:root:Average loss=0.1943549259732931
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2346 | LCC = 0.8619 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.17824578867890897
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2360 | LCC = 0.8643 | SRCC = 0.8632 ]
+INFO:root:Average loss=0.1619350244410527
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2364 | LCC = 0.8641 | SRCC = 0.8635 ]
+INFO:root:Average loss=0.15211974810331297
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2700 | LCC = 0.8614 | SRCC = 0.8646 ]
+INFO:root:Average loss=0.14754558010743216
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2246 | LCC = 0.8660 | SRCC = 0.8653 ]
+INFO:root:Average loss=0.13381555705116346
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2265 | LCC = 0.8643 | SRCC = 0.8636 ]
+INFO:root:Average loss=0.1330194120796827
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2336 | LCC = 0.8609 | SRCC = 0.8598 ]
+INFO:root:Average loss=0.113600281950755
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2329 | LCC = 0.8603 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.12105331923335026
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2298 | LCC = 0.8621 | SRCC = 0.8613 ]
+INFO:root:Average loss=0.10036042824578591
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2263 | LCC = 0.8601 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.10182072384617268
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2314 | LCC = 0.8606 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.09473781125285687
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2310 | LCC = 0.8576 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.08705820329487324
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2319 | LCC = 0.8579 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.08563109654455613
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2329 | LCC = 0.8568 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.07831280372845821
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2314 | LCC = 0.8532 | SRCC = 0.8538 ]
+INFO:root:Average loss=0.07201913056465295
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2296 | LCC = 0.8538 | SRCC = 0.8533 ]
+INFO:root:Average loss=0.06809305189511715
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2373 | LCC = 0.8504 | SRCC = 0.8491 ]
+INFO:root:Average loss=0.06498371594800399
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2267 | LCC = 0.8545 | SRCC = 0.8543 ]
+INFO:root:Average loss=0.06595812155268131
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2296 | LCC = 0.8543 | SRCC = 0.8543 ]
+INFO:root:Average loss=0.06538664802717857
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2268 | LCC = 0.8576 | SRCC = 0.8578 ]
+INFO:root:Average loss=0.05810415176435923
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2497 | LCC = 0.8502 | SRCC = 0.8496 ]
+INFO:root:Average loss=0.05614544608845161
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2337 | LCC = 0.8533 | SRCC = 0.8527 ]
+INFO:root:Average loss=0.05529665674727697
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2328 | LCC = 0.8550 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.05160186551988889
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2337 | LCC = 0.8569 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.05159499486669516
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2387 | LCC = 0.8532 | SRCC = 0.8522 ]
+INFO:root:Average loss=0.05016429894245588
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2413 | LCC = 0.8558 | SRCC = 0.8564 ]
+INFO:root:Average loss=0.049939813713232674
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2285 | LCC = 0.8558 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.045125617478520445
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2283 | LCC = 0.8531 | SRCC = 0.8541 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2412 | LCC = 0.8556 | SRCC = 0.8569 ]

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run3/config.gin

@@ -0,0 +1,31 @@
+# Training config on BVCC dataset.
+
+# ------------------------------
+# Architecture
+# ------------------------------
+ProjectionHead:
+    conv_channels = (32, 32)
+    dense_neurons = (64, 32, 1)
+    use_poolings = (True, True)
+    kernel_size = 5
+    pool_size = 5
+    
+# ------------------------------
+# Dataset
+# ------------------------------
+Bvcc:
+    debug = False
+
+get_dataloader:
+    num_workers = 12
+    shuffle = True
+
+# ------------------------------
+# Training loop
+# ------------------------------
+TrainingLoop:
+    num_epochs = 30
+    learning_rate = 1e-4
+    dataset_cls = @Bvcc
+    weight_decay = 0.0
+    batch_size_train = 64

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run3/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5b8f0ca6c5a820d45f4374ebb07032a93ead16d91abe9103a76d4e177274bd3
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run3/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b30eb8f2e6449acb1d952b4b864ed749e8a51bc83be2e21067ed0a5adf08dc3e
+size 2749638

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run3/train.log

@@ -0,0 +1,99 @@
+INFO:root:Num train speech clips: 4974
+INFO:root:Num val speech clips: 1066
+INFO:root:Num test speech clips: 1066
+INFO:root:Device=cuda
+INFO:root:Average loss=0.4263247901048416
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3341 | LCC = 0.8069 | SRCC = 0.8171 ]
+INFO:root:Average loss=0.25513175225410706
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3166 | LCC = 0.8439 | SRCC = 0.8455 ]
+INFO:root:Average loss=0.2196473790666996
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2450 | LCC = 0.8571 | SRCC = 0.8570 ]
+INFO:root:Average loss=0.2047925625856106
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2476 | LCC = 0.8638 | SRCC = 0.8615 ]
+INFO:root:Average loss=0.18182204749721748
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2253 | LCC = 0.8649 | SRCC = 0.8654 ]
+INFO:root:Average loss=0.16777020138807786
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2687 | LCC = 0.8669 | SRCC = 0.8642 ]
+INFO:root:Average loss=0.1529040923103308
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2244 | LCC = 0.8670 | SRCC = 0.8661 ]
+INFO:root:Average loss=0.1426549872908837
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2271 | LCC = 0.8667 | SRCC = 0.8658 ]
+INFO:root:Average loss=0.13602226256177977
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2409 | LCC = 0.8656 | SRCC = 0.8644 ]
+INFO:root:Average loss=0.12234792648217617
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2595 | LCC = 0.8587 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.1162674955259531
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2263 | LCC = 0.8579 | SRCC = 0.8580 ]
+INFO:root:Average loss=0.11267953671706028
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2359 | LCC = 0.8599 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.10379027775846995
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2269 | LCC = 0.8582 | SRCC = 0.8586 ]
+INFO:root:Average loss=0.09791425835245694
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2255 | LCC = 0.8567 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.09592596732844146
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2295 | LCC = 0.8588 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.08425334879221061
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2298 | LCC = 0.8546 | SRCC = 0.8532 ]
+INFO:root:Average loss=0.08025821952674633
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2456 | LCC = 0.8513 | SRCC = 0.8505 ]
+INFO:root:Average loss=0.08343392835022548
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2326 | LCC = 0.8566 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.07468307653489785
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2361 | LCC = 0.8555 | SRCC = 0.8555 ]
+INFO:root:Average loss=0.06682860679351367
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2438 | LCC = 0.8534 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.06633613244272195
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2390 | LCC = 0.8497 | SRCC = 0.8493 ]
+INFO:root:Average loss=0.07103961238112205
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2335 | LCC = 0.8545 | SRCC = 0.8535 ]
+INFO:root:Average loss=0.06813726669702774
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2272 | LCC = 0.8534 | SRCC = 0.8556 ]
+INFO:root:Average loss=0.05936766902987774
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2390 | LCC = 0.8551 | SRCC = 0.8541 ]
+INFO:root:Average loss=0.0550483016249461
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2504 | LCC = 0.8553 | SRCC = 0.8544 ]
+INFO:root:Average loss=0.05137874006938476
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2397 | LCC = 0.8542 | SRCC = 0.8524 ]
+INFO:root:Average loss=0.05169885684377872
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2353 | LCC = 0.8555 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.04965984797439514
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2455 | LCC = 0.8530 | SRCC = 0.8524 ]
+INFO:root:Average loss=0.0517897374736957
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2654 | LCC = 0.8526 | SRCC = 0.8524 ]
+INFO:root:Average loss=0.04750314476684882
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2903 | LCC = 0.8505 | SRCC = 0.8497 ]
+INFO:root:Average loss=0.046954824756353326
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2455 | LCC = 0.8490 | SRCC = 0.8482 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2395 | LCC = 0.8682 | SRCC = 0.8692 ]

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run4/config.gin

@@ -0,0 +1,31 @@
+# Training config on BVCC dataset.
+
+# ------------------------------
+# Architecture
+# ------------------------------
+ProjectionHead:
+    conv_channels = (32, 32)
+    dense_neurons = (64, 32, 1)
+    use_poolings = (True, True)
+    kernel_size = 5
+    pool_size = 5
+    
+# ------------------------------
+# Dataset
+# ------------------------------
+Bvcc:
+    debug = False
+
+get_dataloader:
+    num_workers = 12
+    shuffle = True
+
+# ------------------------------
+# Training loop
+# ------------------------------
+TrainingLoop:
+    num_epochs = 30
+    learning_rate = 1e-4
+    dataset_cls = @Bvcc
+    weight_decay = 0.0
+    batch_size_train = 64

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run4/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:958fa09fff657fb43010a6c579828c2d6fcda6948a6fb2726f947744c330d27c
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run4/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:056030105fc10ca00c7666e4a380d3fa911042f24e13e3cf928c460560dcb8d2
+size 2749638

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run4/train.log

@@ -0,0 +1,99 @@
+INFO:root:Num train speech clips: 4974
+INFO:root:Num val speech clips: 1066
+INFO:root:Num test speech clips: 1066
+INFO:root:Device=cuda
+INFO:root:Average loss=0.4013510105701593
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3206 | LCC = 0.8193 | SRCC = 0.8251 ]
+INFO:root:Average loss=0.24679987323589814
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2881 | LCC = 0.8489 | SRCC = 0.8498 ]
+INFO:root:Average loss=0.21113414603930253
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2357 | LCC = 0.8600 | SRCC = 0.8607 ]
+INFO:root:Average loss=0.20985345198557928
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2500 | LCC = 0.8619 | SRCC = 0.8615 ]
+INFO:root:Average loss=0.1824611308865058
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2321 | LCC = 0.8650 | SRCC = 0.8638 ]
+INFO:root:Average loss=0.16225871625236976
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2584 | LCC = 0.8605 | SRCC = 0.8629 ]
+INFO:root:Average loss=0.15733267214053717
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2579 | LCC = 0.8644 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.1450038110025418
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2435 | LCC = 0.8663 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.13626796666246194
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2594 | LCC = 0.8638 | SRCC = 0.8644 ]
+INFO:root:Average loss=0.1275765594954674
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2345 | LCC = 0.8638 | SRCC = 0.8635 ]
+INFO:root:Average loss=0.11394043381397541
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2523 | LCC = 0.8602 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.11694239041744134
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2534 | LCC = 0.8609 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.10141534148118435
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2362 | LCC = 0.8613 | SRCC = 0.8615 ]
+INFO:root:Average loss=0.0958498631341335
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2309 | LCC = 0.8592 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.09193842934492306
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2327 | LCC = 0.8610 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.08798203287789455
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2558 | LCC = 0.8586 | SRCC = 0.8589 ]
+INFO:root:Average loss=0.08026020176326618
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2399 | LCC = 0.8574 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.07814991669968152
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2344 | LCC = 0.8567 | SRCC = 0.8575 ]
+INFO:root:Average loss=0.07232317987542886
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2420 | LCC = 0.8578 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.07560246671812657
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2247 | LCC = 0.8610 | SRCC = 0.8617 ]
+INFO:root:Average loss=0.06479993142569676
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2500 | LCC = 0.8510 | SRCC = 0.8524 ]
+INFO:root:Average loss=0.06703913579575527
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2480 | LCC = 0.8539 | SRCC = 0.8553 ]
+INFO:root:Average loss=0.06362319403351882
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2349 | LCC = 0.8557 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.05556169008979431
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2415 | LCC = 0.8553 | SRCC = 0.8568 ]
+INFO:root:Average loss=0.05884094994801741
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2407 | LCC = 0.8554 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.05600188443293938
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2349 | LCC = 0.8561 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.05231332129392868
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2456 | LCC = 0.8539 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.05194630469076145
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2452 | LCC = 0.8524 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.0467600287296451
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2434 | LCC = 0.8525 | SRCC = 0.8538 ]
+INFO:root:Average loss=0.043897231396001116
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2432 | LCC = 0.8491 | SRCC = 0.8513 ]
+INFO:root:Average loss=0.04906416966173893
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2316 | LCC = 0.8545 | SRCC = 0.8568 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2419 | LCC = 0.8703 | SRCC = 0.8689 ]

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run0/config.gin

@@ -0,0 +1,31 @@
+# Training config on BVCC dataset.
+
+# ------------------------------
+# Architecture
+# ------------------------------
+ProjectionHead:
+    conv_channels = (32, 32)
+    dense_neurons = (64, 32, 1)
+    use_poolings = (True, True)
+    kernel_size = 5
+    pool_size = 5
+    
+# ------------------------------
+# Dataset
+# ------------------------------
+Bvcc:
+    debug = False
+
+get_dataloader:
+    num_workers = 12
+    shuffle = True
+
+# ------------------------------
+# Training loop
+# ------------------------------
+TrainingLoop:
+    num_epochs = 30
+    learning_rate = 1e-4
+    dataset_cls = @Bvcc
+    weight_decay = 0.0
+    batch_size_train = 64

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run0/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc09f975ea570dd8de6495633bb41e28b4c6d4f18566a8ac286eab71406c7859
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run0/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31bfb306a992a5e589efc0d6caaf910a5fca212e947669a6021791ae35b79c56
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run0/train.log

@@ -0,0 +1,99 @@
+INFO:root:Num train speech clips: 4974
+INFO:root:Num val speech clips: 1066
+INFO:root:Num test speech clips: 1066
+INFO:root:Device=cuda
+INFO:root:Average loss=0.3647264561209923
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3369 | LCC = 0.8328 | SRCC = 0.8340 ]
+INFO:root:Average loss=0.2450862244153634
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2823 | LCC = 0.8550 | SRCC = 0.8510 ]
+INFO:root:Average loss=0.22131552088719147
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2811 | LCC = 0.8607 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.20576652865379286
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2710 | LCC = 0.8626 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.18918423899091208
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2542 | LCC = 0.8639 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.17924471505177328
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2549 | LCC = 0.8651 | SRCC = 0.8648 ]
+INFO:root:Average loss=0.1597173732633774
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2493 | LCC = 0.8634 | SRCC = 0.8607 ]
+INFO:root:Average loss=0.15701985082183129
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2403 | LCC = 0.8630 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.14052439079834864
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2499 | LCC = 0.8616 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.12632038396520492
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2446 | LCC = 0.8585 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.11942222905464661
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2648 | LCC = 0.8558 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.11359471928041714
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2774 | LCC = 0.8484 | SRCC = 0.8484 ]
+INFO:root:Average loss=0.10948953042045617
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2499 | LCC = 0.8518 | SRCC = 0.8521 ]
+INFO:root:Average loss=0.10058757051443443
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2358 | LCC = 0.8559 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.09444239783363464
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2476 | LCC = 0.8510 | SRCC = 0.8504 ]
+INFO:root:Average loss=0.08680354794248557
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2495 | LCC = 0.8524 | SRCC = 0.8513 ]
+INFO:root:Average loss=0.08416341503079121
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2892 | LCC = 0.8487 | SRCC = 0.8502 ]
+INFO:root:Average loss=0.07809728952363515
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2460 | LCC = 0.8490 | SRCC = 0.8485 ]
+INFO:root:Average loss=0.0781081745830866
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2471 | LCC = 0.8498 | SRCC = 0.8490 ]
+INFO:root:Average loss=0.07456183290252319
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2315 | LCC = 0.8516 | SRCC = 0.8520 ]
+INFO:root:Average loss=0.06630134319838805
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2422 | LCC = 0.8489 | SRCC = 0.8490 ]
+INFO:root:Average loss=0.06708064737419288
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2452 | LCC = 0.8478 | SRCC = 0.8481 ]
+INFO:root:Average loss=0.06410350593236777
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2479 | LCC = 0.8456 | SRCC = 0.8446 ]
+INFO:root:Average loss=0.060496386809226796
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2477 | LCC = 0.8476 | SRCC = 0.8480 ]
+INFO:root:Average loss=0.06087472643225621
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2350 | LCC = 0.8523 | SRCC = 0.8517 ]
+INFO:root:Average loss=0.05840050056576729
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2441 | LCC = 0.8481 | SRCC = 0.8484 ]
+INFO:root:Average loss=0.05399366649679649
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2451 | LCC = 0.8481 | SRCC = 0.8485 ]
+INFO:root:Average loss=0.05104239884381875
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2514 | LCC = 0.8447 | SRCC = 0.8454 ]
+INFO:root:Average loss=0.046379657390599065
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2477 | LCC = 0.8463 | SRCC = 0.8453 ]
+INFO:root:Average loss=0.04524605465718569
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2438 | LCC = 0.8480 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.049099207640840456
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2414 | LCC = 0.8479 | SRCC = 0.8470 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2496 | LCC = 0.8601 | SRCC = 0.8602 ]

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run1/config.gin

@@ -0,0 +1,31 @@
+# Training config on BVCC dataset.
+
+# ------------------------------
+# Architecture
+# ------------------------------
+ProjectionHead:
+    conv_channels = (32, 32)
+    dense_neurons = (64, 32, 1)
+    use_poolings = (True, True)
+    kernel_size = 5
+    pool_size = 5
+    
+# ------------------------------
+# Dataset
+# ------------------------------
+Bvcc:
+    debug = False
+
+get_dataloader:
+    num_workers = 12
+    shuffle = True
+
+# ------------------------------
+# Training loop
+# ------------------------------
+TrainingLoop:
+    num_epochs = 30
+    learning_rate = 1e-4
+    dataset_cls = @Bvcc
+    weight_decay = 0.0
+    batch_size_train = 64

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run1/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6edcd92e8f241b1f39f8abbe2248e38b3de0dc1ee4f8ca2bc0a5b86cc3510211
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run1/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79b99475319d9492f9f4f13909f52a48f31a58cdfeedc1a7abb06b0969bd5ebf
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run1/train.log

@@ -0,0 +1,99 @@
+INFO:root:Num train speech clips: 4974
+INFO:root:Num val speech clips: 1066
+INFO:root:Num test speech clips: 1066
+INFO:root:Device=cuda
+INFO:root:Average loss=0.36341217427681655
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3052 | LCC = 0.8347 | SRCC = 0.8364 ]
+INFO:root:Average loss=0.2456711721725953
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2770 | LCC = 0.8508 | SRCC = 0.8501 ]
+INFO:root:Average loss=0.22876097242801618
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2806 | LCC = 0.8600 | SRCC = 0.8571 ]
+INFO:root:Average loss=0.19449595438364226
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2520 | LCC = 0.8658 | SRCC = 0.8634 ]
+INFO:root:Average loss=0.18043607282332885
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2403 | LCC = 0.8666 | SRCC = 0.8662 ]
+INFO:root:Average loss=0.16845933796885687
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2698 | LCC = 0.8684 | SRCC = 0.8658 ]
+INFO:root:Average loss=0.15109611665591216
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2343 | LCC = 0.8631 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.14470888979924032
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3042 | LCC = 0.8581 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.13583920628596574
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2527 | LCC = 0.8610 | SRCC = 0.8597 ]
+INFO:root:Average loss=0.12527909263586387
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2546 | LCC = 0.8603 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.11729279284675916
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2624 | LCC = 0.8578 | SRCC = 0.8570 ]
+INFO:root:Average loss=0.11554487584493099
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2687 | LCC = 0.8585 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.1017474483889647
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2518 | LCC = 0.8550 | SRCC = 0.8552 ]
+INFO:root:Average loss=0.09209378579488167
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2672 | LCC = 0.8570 | SRCC = 0.8562 ]
+INFO:root:Average loss=0.0961250891097081
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2327 | LCC = 0.8548 | SRCC = 0.8546 ]
+INFO:root:Average loss=0.08398169837892056
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2401 | LCC = 0.8552 | SRCC = 0.8550 ]
+INFO:root:Average loss=0.08130157533555459
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2373 | LCC = 0.8544 | SRCC = 0.8550 ]
+INFO:root:Average loss=0.07885749332415752
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2506 | LCC = 0.8521 | SRCC = 0.8507 ]
+INFO:root:Average loss=0.0726623892879639
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2421 | LCC = 0.8506 | SRCC = 0.8506 ]
+INFO:root:Average loss=0.067769757591379
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2533 | LCC = 0.8547 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.06732550139228503
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2397 | LCC = 0.8534 | SRCC = 0.8527 ]
+INFO:root:Average loss=0.06061897364755472
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2433 | LCC = 0.8513 | SRCC = 0.8500 ]
+INFO:root:Average loss=0.05890513985202862
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2651 | LCC = 0.8465 | SRCC = 0.8460 ]
+INFO:root:Average loss=0.05775275702277819
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2444 | LCC = 0.8517 | SRCC = 0.8512 ]
+INFO:root:Average loss=0.05580917048530701
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2351 | LCC = 0.8527 | SRCC = 0.8524 ]
+INFO:root:Average loss=0.054117803437969625
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2570 | LCC = 0.8521 | SRCC = 0.8511 ]
+INFO:root:Average loss=0.05272514511568424
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2490 | LCC = 0.8512 | SRCC = 0.8503 ]
+INFO:root:Average loss=0.05023255418890562
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2417 | LCC = 0.8495 | SRCC = 0.8488 ]
+INFO:root:Average loss=0.048341174490558796
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2457 | LCC = 0.8500 | SRCC = 0.8482 ]
+INFO:root:Average loss=0.05171361782898506
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2507 | LCC = 0.8501 | SRCC = 0.8475 ]
+INFO:root:Average loss=0.04439792696100015
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2393 | LCC = 0.8525 | SRCC = 0.8513 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2752 | LCC = 0.8590 | SRCC = 0.8597 ]

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run2/config.gin

@@ -0,0 +1,31 @@
+# Training config on BVCC dataset.
+
+# ------------------------------
+# Architecture
+# ------------------------------
+ProjectionHead:
+    conv_channels = (32, 32)
+    dense_neurons = (64, 32, 1)
+    use_poolings = (True, True)
+    kernel_size = 5
+    pool_size = 5
+    
+# ------------------------------
+# Dataset
+# ------------------------------
+Bvcc:
+    debug = False
+
+get_dataloader:
+    num_workers = 12
+    shuffle = True
+
+# ------------------------------
+# Training loop
+# ------------------------------
+TrainingLoop:
+    num_epochs = 30
+    learning_rate = 1e-4
+    dataset_cls = @Bvcc
+    weight_decay = 0.0
+    batch_size_train = 64

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run2/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08e8b0573262bd617b63ee395cd37d97ec4b24f3636ccf53738d2f2d81ffafdf
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run2/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be62dfaba70922177033b2150da16a5e3ab228354055eaeec3db2d21060be666
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run2/train.log

@@ -0,0 +1,99 @@
+INFO:root:Num train speech clips: 4974
+INFO:root:Num val speech clips: 1066
+INFO:root:Num test speech clips: 1066
+INFO:root:Device=cuda
+INFO:root:Average loss=0.3910743683958665
+INFO:root:
+[Valid][0][UTT][ MSE = 0.2994 | LCC = 0.8279 | SRCC = 0.8289 ]
+INFO:root:Average loss=0.2482904673386843
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3127 | LCC = 0.8494 | SRCC = 0.8471 ]
+INFO:root:Average loss=0.22257250165328002
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2769 | LCC = 0.8563 | SRCC = 0.8557 ]
+INFO:root:Average loss=0.20747650806338358
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2727 | LCC = 0.8606 | SRCC = 0.8617 ]
+INFO:root:Average loss=0.1903773407714489
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2460 | LCC = 0.8674 | SRCC = 0.8654 ]
+INFO:root:Average loss=0.1798965913745073
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2297 | LCC = 0.8658 | SRCC = 0.8654 ]
+INFO:root:Average loss=0.16481705525746712
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2283 | LCC = 0.8657 | SRCC = 0.8646 ]
+INFO:root:Average loss=0.15823762195232588
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2425 | LCC = 0.8645 | SRCC = 0.8646 ]
+INFO:root:Average loss=0.14712723774405626
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2633 | LCC = 0.8622 | SRCC = 0.8617 ]
+INFO:root:Average loss=0.137577229203322
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2257 | LCC = 0.8612 | SRCC = 0.8609 ]
+INFO:root:Average loss=0.12150792701122089
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2346 | LCC = 0.8558 | SRCC = 0.8553 ]
+INFO:root:Average loss=0.11626947489686501
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2365 | LCC = 0.8563 | SRCC = 0.8569 ]
+INFO:root:Average loss=0.11079482380778362
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2468 | LCC = 0.8491 | SRCC = 0.8498 ]
+INFO:root:Average loss=0.10233893474707237
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2654 | LCC = 0.8504 | SRCC = 0.8500 ]
+INFO:root:Average loss=0.09710497638353935
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2451 | LCC = 0.8523 | SRCC = 0.8514 ]
+INFO:root:Average loss=0.08908081900041837
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2518 | LCC = 0.8505 | SRCC = 0.8510 ]
+INFO:root:Average loss=0.08982657645948422
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2517 | LCC = 0.8472 | SRCC = 0.8478 ]
+INFO:root:Average loss=0.08223194442689419
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2398 | LCC = 0.8512 | SRCC = 0.8511 ]
+INFO:root:Average loss=0.08082248733784908
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2622 | LCC = 0.8478 | SRCC = 0.8479 ]
+INFO:root:Average loss=0.07176543657596295
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2849 | LCC = 0.8428 | SRCC = 0.8425 ]
+INFO:root:Average loss=0.07014165833019294
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2666 | LCC = 0.8426 | SRCC = 0.8417 ]
+INFO:root:Average loss=0.067525883563436
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2464 | LCC = 0.8464 | SRCC = 0.8450 ]
+INFO:root:Average loss=0.06632072597933121
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2489 | LCC = 0.8459 | SRCC = 0.8463 ]
+INFO:root:Average loss=0.07072250764721479
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2553 | LCC = 0.8481 | SRCC = 0.8466 ]
+INFO:root:Average loss=0.05845999961289076
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2555 | LCC = 0.8432 | SRCC = 0.8416 ]
+INFO:root:Average loss=0.05378421171544454
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2618 | LCC = 0.8441 | SRCC = 0.8429 ]
+INFO:root:Average loss=0.053914408366649576
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2450 | LCC = 0.8444 | SRCC = 0.8448 ]
+INFO:root:Average loss=0.05011503995420077
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2563 | LCC = 0.8434 | SRCC = 0.8427 ]
+INFO:root:Average loss=0.05067756752937268
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2394 | LCC = 0.8503 | SRCC = 0.8510 ]
+INFO:root:Average loss=0.05002134388838059
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2468 | LCC = 0.8464 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.04614516285558542
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2442 | LCC = 0.8477 | SRCC = 0.8477 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2500 | LCC = 0.8586 | SRCC = 0.8583 ]

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run3/config.gin

@@ -0,0 +1,31 @@
+# Training config on BVCC dataset.
+
+# ------------------------------
+# Architecture
+# ------------------------------
+ProjectionHead:
+    conv_channels = (32, 32)
+    dense_neurons = (64, 32, 1)
+    use_poolings = (True, True)
+    kernel_size = 5
+    pool_size = 5
+    
+# ------------------------------
+# Dataset
+# ------------------------------
+Bvcc:
+    debug = False
+
+get_dataloader:
+    num_workers = 12
+    shuffle = True
+
+# ------------------------------
+# Training loop
+# ------------------------------
+TrainingLoop:
+    num_epochs = 30
+    learning_rate = 1e-4
+    dataset_cls = @Bvcc
+    weight_decay = 0.0
+    batch_size_train = 64

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run3/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:032cda5e208ec0c92aba7fd2a684d435143dc5cd174d8d263ec202233a3fb92d
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run3/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:142b530226ced77155d0639b58fa5692e277369bb6c04e512e1ffb9c8f21e3e8
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run3/train.log

@@ -0,0 +1,99 @@
+INFO:root:Num train speech clips: 4974
+INFO:root:Num val speech clips: 1066
+INFO:root:Num test speech clips: 1066
+INFO:root:Device=cuda
+INFO:root:Average loss=0.45257461338471144
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3301 | LCC = 0.8254 | SRCC = 0.8307 ]
+INFO:root:Average loss=0.24987471543061426
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2638 | LCC = 0.8459 | SRCC = 0.8506 ]
+INFO:root:Average loss=0.21798543517406171
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2613 | LCC = 0.8593 | SRCC = 0.8598 ]
+INFO:root:Average loss=0.19310585953868353
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2482 | LCC = 0.8644 | SRCC = 0.8641 ]
+INFO:root:Average loss=0.17893173593359116
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2462 | LCC = 0.8615 | SRCC = 0.8588 ]
+INFO:root:Average loss=0.1604620206814546
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2317 | LCC = 0.8629 | SRCC = 0.8613 ]
+INFO:root:Average loss=0.14410021853370544
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2373 | LCC = 0.8595 | SRCC = 0.8571 ]
+INFO:root:Average loss=0.1442149166877453
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2284 | LCC = 0.8592 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.13009902242666635
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2336 | LCC = 0.8574 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.12277384408009358
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2321 | LCC = 0.8586 | SRCC = 0.8586 ]
+INFO:root:Average loss=0.1101332270564177
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2406 | LCC = 0.8533 | SRCC = 0.8537 ]
+INFO:root:Average loss=0.1030655565361182
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2380 | LCC = 0.8528 | SRCC = 0.8520 ]
+INFO:root:Average loss=0.10142424148626816
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2334 | LCC = 0.8537 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.0928691405420884
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2425 | LCC = 0.8513 | SRCC = 0.8509 ]
+INFO:root:Average loss=0.08392868463236552
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2499 | LCC = 0.8502 | SRCC = 0.8493 ]
+INFO:root:Average loss=0.07562753782631496
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2454 | LCC = 0.8471 | SRCC = 0.8459 ]
+INFO:root:Average loss=0.07394624014313404
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2429 | LCC = 0.8473 | SRCC = 0.8470 ]
+INFO:root:Average loss=0.06798708797074281
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2501 | LCC = 0.8445 | SRCC = 0.8442 ]
+INFO:root:Average loss=0.06676959738326378
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2391 | LCC = 0.8488 | SRCC = 0.8476 ]
+INFO:root:Average loss=0.06508324992580292
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2480 | LCC = 0.8432 | SRCC = 0.8424 ]
+INFO:root:Average loss=0.061816331811058216
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2513 | LCC = 0.8446 | SRCC = 0.8443 ]
+INFO:root:Average loss=0.060355358160077
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2285 | LCC = 0.8512 | SRCC = 0.8527 ]
+INFO:root:Average loss=0.05913594076171135
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2445 | LCC = 0.8465 | SRCC = 0.8452 ]
+INFO:root:Average loss=0.05481694161127775
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2351 | LCC = 0.8480 | SRCC = 0.8471 ]
+INFO:root:Average loss=0.05041407796148306
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2406 | LCC = 0.8464 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.05042444432202058
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2393 | LCC = 0.8465 | SRCC = 0.8461 ]
+INFO:root:Average loss=0.04690109766446627
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2425 | LCC = 0.8454 | SRCC = 0.8454 ]
+INFO:root:Average loss=0.04716737792851069
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2463 | LCC = 0.8431 | SRCC = 0.8430 ]
+INFO:root:Average loss=0.0466637581300277
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2567 | LCC = 0.8460 | SRCC = 0.8493 ]
+INFO:root:Average loss=0.041203983271351226
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2431 | LCC = 0.8455 | SRCC = 0.8452 ]
+INFO:root:Average loss=0.04254784334737521
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2443 | LCC = 0.8445 | SRCC = 0.8454 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2443 | LCC = 0.8644 | SRCC = 0.8621 ]

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run4/config.gin

@@ -0,0 +1,31 @@
+# Training config on BVCC dataset.
+
+# ------------------------------
+# Architecture
+# ------------------------------
+ProjectionHead:
+    conv_channels = (32, 32)
+    dense_neurons = (64, 32, 1)
+    use_poolings = (True, True)
+    kernel_size = 5
+    pool_size = 5
+    
+# ------------------------------
+# Dataset
+# ------------------------------
+Bvcc:
+    debug = False
+
+get_dataloader:
+    num_workers = 12
+    shuffle = True
+
+# ------------------------------
+# Training loop
+# ------------------------------
+TrainingLoop:
+    num_epochs = 30
+    learning_rate = 1e-4
+    dataset_cls = @Bvcc
+    weight_decay = 0.0
+    batch_size_train = 64

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run4/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4a861c4876b0d32e8af1f90ddfb8c8b087c413c50c70792d90fbc02309ef78a
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run4/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6e4106e0b829ed7399b5bbc4d1290e10a6806cf25cdf6e2b422952a933aa0f6
+size 2749702