Upload SSL Layer MOS checkpoints (part 8)

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer5/run2/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6b50c9387dcbe58e205f9a57677b09388adccb43bdcedee1f59d0478f4c8355
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer5/run2/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b99ff50d28a48dacfe0bf472bec9104893ed3d08d396214babc54bebba8b268b
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/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=0.3602916034750449
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3017 | LCC = 0.8297 | SRCC = 0.8342 ]
+INFO:root:Average loss=0.22860305699018332
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2704 | LCC = 0.8552 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.2008502225463207
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2308 | LCC = 0.8624 | SRCC = 0.8625 ]
+INFO:root:Average loss=0.185303858648508
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2760 | LCC = 0.8650 | SRCC = 0.8637 ]
+INFO:root:Average loss=0.17254548395673433
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2495 | LCC = 0.8698 | SRCC = 0.8676 ]
+INFO:root:Average loss=0.15667101721733045
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2468 | LCC = 0.8667 | SRCC = 0.8644 ]
+INFO:root:Average loss=0.1475692986486814
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2355 | LCC = 0.8645 | SRCC = 0.8632 ]
+INFO:root:Average loss=0.13897179391903755
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2490 | LCC = 0.8602 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.1250226007631192
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2292 | LCC = 0.8599 | SRCC = 0.8590 ]
+INFO:root:Average loss=0.1178060746154724
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2492 | LCC = 0.8604 | SRCC = 0.8591 ]
+INFO:root:Average loss=0.10731051575679046
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2352 | LCC = 0.8585 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.10309349201046504
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2484 | LCC = 0.8597 | SRCC = 0.8569 ]
+INFO:root:Average loss=0.09537047563264003
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2379 | LCC = 0.8544 | SRCC = 0.8522 ]
+INFO:root:Average loss=0.08652413985094963
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2674 | LCC = 0.8531 | SRCC = 0.8496 ]
+INFO:root:Average loss=0.0837886471969959
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2418 | LCC = 0.8534 | SRCC = 0.8531 ]
+INFO:root:Average loss=0.07590657128737523
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2412 | LCC = 0.8523 | SRCC = 0.8515 ]
+INFO:root:Average loss=0.07344822246485795
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2441 | LCC = 0.8499 | SRCC = 0.8483 ]
+INFO:root:Average loss=0.07031746958501828
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2497 | LCC = 0.8503 | SRCC = 0.8488 ]
+INFO:root:Average loss=0.0652227365435698
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2388 | LCC = 0.8523 | SRCC = 0.8509 ]
+INFO:root:Average loss=0.06416319434841473
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2686 | LCC = 0.8484 | SRCC = 0.8486 ]
+INFO:root:Average loss=0.06138015863222954
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2499 | LCC = 0.8480 | SRCC = 0.8456 ]
+INFO:root:Average loss=0.05661991114417712
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2570 | LCC = 0.8495 | SRCC = 0.8489 ]
+INFO:root:Average loss=0.05485928946962723
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2561 | LCC = 0.8467 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.05111394521708672
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2419 | LCC = 0.8487 | SRCC = 0.8462 ]
+INFO:root:Average loss=0.05058705133314316
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2489 | LCC = 0.8469 | SRCC = 0.8445 ]
+INFO:root:Average loss=0.04935907319379158
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2400 | LCC = 0.8457 | SRCC = 0.8447 ]
+INFO:root:Average loss=0.04663730957187139
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2469 | LCC = 0.8478 | SRCC = 0.8456 ]
+INFO:root:Average loss=0.04633009780007295
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2476 | LCC = 0.8467 | SRCC = 0.8461 ]
+INFO:root:Average loss=0.04485997715248512
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2528 | LCC = 0.8471 | SRCC = 0.8462 ]
+INFO:root:Average loss=0.04893341567367315
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2455 | LCC = 0.8453 | SRCC = 0.8441 ]
+INFO:root:Average loss=0.04172439204576688
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2377 | LCC = 0.8469 | SRCC = 0.8457 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2426 | LCC = 0.8652 | SRCC = 0.8651 ]

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/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_8epoch/layer5/run3/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69884306410746ecb8677bfefed6eb0b64c36c6dc02b6daca2cad34720312c19
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer5/run3/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4bc767021def3f224f03464651aad93bce7804f579a71f1711116c8bafb7ab1a
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/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=0.362190163288361
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3009 | LCC = 0.8301 | SRCC = 0.8381 ]
+INFO:root:Average loss=0.2284203271071116
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2673 | LCC = 0.8539 | SRCC = 0.8531 ]
+INFO:root:Average loss=0.20273739557999831
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2613 | LCC = 0.8658 | SRCC = 0.8636 ]
+INFO:root:Average loss=0.18265455999435523
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2410 | LCC = 0.8670 | SRCC = 0.8649 ]
+INFO:root:Average loss=0.17725740506862983
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2656 | LCC = 0.8689 | SRCC = 0.8669 ]
+INFO:root:Average loss=0.1582477321036351
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2515 | LCC = 0.8697 | SRCC = 0.8684 ]
+INFO:root:Average loss=0.14821337889402342
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2149 | LCC = 0.8709 | SRCC = 0.8709 ]
+INFO:root:Average loss=0.14272961555383143
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2318 | LCC = 0.8722 | SRCC = 0.8715 ]
+INFO:root:Average loss=0.12586095403784361
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2161 | LCC = 0.8669 | SRCC = 0.8674 ]
+INFO:root:Average loss=0.12062422826122014
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2265 | LCC = 0.8707 | SRCC = 0.8704 ]
+INFO:root:Average loss=0.11422893978082217
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2333 | LCC = 0.8650 | SRCC = 0.8649 ]
+INFO:root:Average loss=0.10612992885021064
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2235 | LCC = 0.8654 | SRCC = 0.8653 ]
+INFO:root:Average loss=0.09691693519170468
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2206 | LCC = 0.8645 | SRCC = 0.8653 ]
+INFO:root:Average loss=0.09305107727264747
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2303 | LCC = 0.8633 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.0835542368869751
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2274 | LCC = 0.8627 | SRCC = 0.8636 ]
+INFO:root:Average loss=0.08460547426381172
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2450 | LCC = 0.8617 | SRCC = 0.8615 ]
+INFO:root:Average loss=0.07684859504493383
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2259 | LCC = 0.8592 | SRCC = 0.8591 ]
+INFO:root:Average loss=0.07778260527321926
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2414 | LCC = 0.8588 | SRCC = 0.8589 ]
+INFO:root:Average loss=0.06990714949101974
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2353 | LCC = 0.8605 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.06567355474600425
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2199 | LCC = 0.8611 | SRCC = 0.8613 ]
+INFO:root:Average loss=0.06059241590973658
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2355 | LCC = 0.8560 | SRCC = 0.8553 ]
+INFO:root:Average loss=0.058565773404179476
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2288 | LCC = 0.8588 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.05620270083921078
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2361 | LCC = 0.8612 | SRCC = 0.8620 ]
+INFO:root:Average loss=0.05402580209267445
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2293 | LCC = 0.8571 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.0521805359958074
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2434 | LCC = 0.8537 | SRCC = 0.8538 ]
+INFO:root:Average loss=0.0494673982835733
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2347 | LCC = 0.8553 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.047855974055635624
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2351 | LCC = 0.8556 | SRCC = 0.8555 ]
+INFO:root:Average loss=0.045182778691061035
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2422 | LCC = 0.8575 | SRCC = 0.8570 ]
+INFO:root:Average loss=0.04461886026920416
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2312 | LCC = 0.8586 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.04863246611486643
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2319 | LCC = 0.8594 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.04162825722820484
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2430 | LCC = 0.8544 | SRCC = 0.8547 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2513 | LCC = 0.8650 | SRCC = 0.8636 ]

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/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_8epoch/layer5/run4/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc0bad2deb9bb1684da3dabc3f58b3b06de9ec6bf59170afbc4d9e27cf501f80
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer5/run4/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:532c95da86fd5cb69f939eae1706a57c03831d7bf58fbf91b39978c90f6bbfd0
+size 2749638

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer5/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=0.42477776454045224
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3218 | LCC = 0.8238 | SRCC = 0.8294 ]
+INFO:root:Average loss=0.23074349073263314
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2693 | LCC = 0.8502 | SRCC = 0.8498 ]
+INFO:root:Average loss=0.21214958547781676
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2734 | LCC = 0.8590 | SRCC = 0.8568 ]
+INFO:root:Average loss=0.18804034265952232
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2621 | LCC = 0.8589 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.16857376904823843
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2477 | LCC = 0.8623 | SRCC = 0.8611 ]
+INFO:root:Average loss=0.15654911922338682
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2614 | LCC = 0.8590 | SRCC = 0.8580 ]
+INFO:root:Average loss=0.144945139113145
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2632 | LCC = 0.8569 | SRCC = 0.8559 ]
+INFO:root:Average loss=0.13225655763959274
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2594 | LCC = 0.8559 | SRCC = 0.8553 ]
+INFO:root:Average loss=0.1209458742195215
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2373 | LCC = 0.8545 | SRCC = 0.8548 ]
+INFO:root:Average loss=0.11163341435484397
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2365 | LCC = 0.8569 | SRCC = 0.8569 ]
+INFO:root:Average loss=0.10707223711487575
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2548 | LCC = 0.8537 | SRCC = 0.8537 ]
+INFO:root:Average loss=0.09564200382775222
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2528 | LCC = 0.8518 | SRCC = 0.8520 ]
+INFO:root:Average loss=0.09240232288646393
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2525 | LCC = 0.8513 | SRCC = 0.8511 ]
+INFO:root:Average loss=0.08303699680627921
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2726 | LCC = 0.8468 | SRCC = 0.8463 ]
+INFO:root:Average loss=0.08223164812303506
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2445 | LCC = 0.8524 | SRCC = 0.8526 ]
+INFO:root:Average loss=0.07442832212799634
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2502 | LCC = 0.8453 | SRCC = 0.8476 ]
+INFO:root:Average loss=0.07438317820047721
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2584 | LCC = 0.8472 | SRCC = 0.8471 ]
+INFO:root:Average loss=0.06566222432332161
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2409 | LCC = 0.8496 | SRCC = 0.8500 ]
+INFO:root:Average loss=0.0633346375842125
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2424 | LCC = 0.8464 | SRCC = 0.8473 ]
+INFO:root:Average loss=0.062345233507072315
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2516 | LCC = 0.8499 | SRCC = 0.8480 ]
+INFO:root:Average loss=0.06149311549961567
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2590 | LCC = 0.8481 | SRCC = 0.8481 ]
+INFO:root:Average loss=0.057354543883448995
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2502 | LCC = 0.8477 | SRCC = 0.8473 ]
+INFO:root:Average loss=0.05565970307454849
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2667 | LCC = 0.8479 | SRCC = 0.8477 ]
+INFO:root:Average loss=0.05516067046958666
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2483 | LCC = 0.8464 | SRCC = 0.8475 ]
+INFO:root:Average loss=0.04986106446729256
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2603 | LCC = 0.8465 | SRCC = 0.8480 ]
+INFO:root:Average loss=0.050457364425827295
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2401 | LCC = 0.8512 | SRCC = 0.8521 ]
+INFO:root:Average loss=0.04614992647503431
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2470 | LCC = 0.8502 | SRCC = 0.8497 ]
+INFO:root:Average loss=0.04329039013156524
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2521 | LCC = 0.8483 | SRCC = 0.8483 ]
+INFO:root:Average loss=0.0468870301802571
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2459 | LCC = 0.8471 | SRCC = 0.8469 ]
+INFO:root:Average loss=0.041542455864449344
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2544 | LCC = 0.8480 | SRCC = 0.8482 ]
+INFO:root:Average loss=0.04111270451297363
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2684 | LCC = 0.8404 | SRCC = 0.8402 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2458 | LCC = 0.8641 | SRCC = 0.8660 ]

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/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_8epoch/layer6/run0/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4ce8cbfd7b297a9bc73cf79d8557e7f9b95f747a18d6c73a5b0c34fd755f043
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/run0/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9fc80e491c15e22e6525277680d7e91ea160cb9d89a4e88013c7bc5c8dafb32b
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/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=0.37467708037449765
+INFO:root:
+[Valid][0][UTT][ MSE = 0.2990 | LCC = 0.8281 | SRCC = 0.8343 ]
+INFO:root:Average loss=0.2333670946267935
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3143 | LCC = 0.8528 | SRCC = 0.8531 ]
+INFO:root:Average loss=0.20628967842994592
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2688 | LCC = 0.8577 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.18531899908796334
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2608 | LCC = 0.8651 | SRCC = 0.8653 ]
+INFO:root:Average loss=0.17028685458577597
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2490 | LCC = 0.8692 | SRCC = 0.8693 ]
+INFO:root:Average loss=0.15651398706130493
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2452 | LCC = 0.8712 | SRCC = 0.8696 ]
+INFO:root:Average loss=0.14767660764165413
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2434 | LCC = 0.8708 | SRCC = 0.8693 ]
+INFO:root:Average loss=0.1454870343590394
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2332 | LCC = 0.8654 | SRCC = 0.8666 ]
+INFO:root:Average loss=0.13086320966100082
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2651 | LCC = 0.8646 | SRCC = 0.8659 ]
+INFO:root:Average loss=0.13209669798230514
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2353 | LCC = 0.8653 | SRCC = 0.8648 ]
+INFO:root:Average loss=0.11594731981555621
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2512 | LCC = 0.8638 | SRCC = 0.8637 ]
+INFO:root:Average loss=0.10751021524461415
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2630 | LCC = 0.8627 | SRCC = 0.8650 ]
+INFO:root:Average loss=0.10847543781766525
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2342 | LCC = 0.8638 | SRCC = 0.8646 ]
+INFO:root:Average loss=0.09495962396837197
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2558 | LCC = 0.8581 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.09150164330808017
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2514 | LCC = 0.8579 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.08520652950765231
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2599 | LCC = 0.8562 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.07926359218664658
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2289 | LCC = 0.8552 | SRCC = 0.8571 ]
+INFO:root:Average loss=0.07587479799985886
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2405 | LCC = 0.8530 | SRCC = 0.8549 ]
+INFO:root:Average loss=0.07495530842779538
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2702 | LCC = 0.8545 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.06786195890834698
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2363 | LCC = 0.8514 | SRCC = 0.8527 ]
+INFO:root:Average loss=0.06390005674881813
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2880 | LCC = 0.8501 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.06459033708923902
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2764 | LCC = 0.8492 | SRCC = 0.8524 ]
+INFO:root:Average loss=0.0626679030366433
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2336 | LCC = 0.8538 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.057929873609772094
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2522 | LCC = 0.8498 | SRCC = 0.8517 ]
+INFO:root:Average loss=0.054576437609891094
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2479 | LCC = 0.8518 | SRCC = 0.8535 ]
+INFO:root:Average loss=0.05767752265987488
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2480 | LCC = 0.8496 | SRCC = 0.8512 ]
+INFO:root:Average loss=0.05206704005981103
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2765 | LCC = 0.8493 | SRCC = 0.8518 ]
+INFO:root:Average loss=0.04931540763339935
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2568 | LCC = 0.8494 | SRCC = 0.8503 ]
+INFO:root:Average loss=0.04757789340920937
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2551 | LCC = 0.8521 | SRCC = 0.8531 ]
+INFO:root:Average loss=0.04776382002119835
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2409 | LCC = 0.8512 | SRCC = 0.8522 ]
+INFO:root:Average loss=0.045480646479588285
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2519 | LCC = 0.8506 | SRCC = 0.8530 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2367 | LCC = 0.8643 | SRCC = 0.8639 ]

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/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_8epoch/layer6/run1/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4da41d337bab409df9120bda5f858c3242ee8a66301e702c145e23f333a0aa28
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/run1/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56b570322548468a0e5d392aae1ef40f04c7c330bd8bbc864435d326f85340d6
+size 2749638

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/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=0.3797180706109756
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3046 | LCC = 0.8346 | SRCC = 0.8365 ]
+INFO:root:Average loss=0.2305016590234561
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2970 | LCC = 0.8560 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.21048951024810472
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2487 | LCC = 0.8620 | SRCC = 0.8611 ]
+INFO:root:Average loss=0.1981869487044139
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2646 | LCC = 0.8666 | SRCC = 0.8650 ]
+INFO:root:Average loss=0.18076091221509835
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2333 | LCC = 0.8649 | SRCC = 0.8658 ]
+INFO:root:Average loss=0.16797982777158418
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2682 | LCC = 0.8642 | SRCC = 0.8651 ]
+INFO:root:Average loss=0.1563210419546335
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2478 | LCC = 0.8661 | SRCC = 0.8650 ]
+INFO:root:Average loss=0.1436955794118918
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2476 | LCC = 0.8625 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.1337132291534008
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2725 | LCC = 0.8544 | SRCC = 0.8546 ]
+INFO:root:Average loss=0.13001818515551397
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2334 | LCC = 0.8595 | SRCC = 0.8588 ]
+INFO:root:Average loss=0.11326501470727798
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2451 | LCC = 0.8579 | SRCC = 0.8586 ]
+INFO:root:Average loss=0.10822764717233486
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2424 | LCC = 0.8543 | SRCC = 0.8550 ]
+INFO:root:Average loss=0.10454240384010169
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2552 | LCC = 0.8536 | SRCC = 0.8525 ]
+INFO:root:Average loss=0.09520155578278579
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2694 | LCC = 0.8526 | SRCC = 0.8515 ]
+INFO:root:Average loss=0.08673519899065678
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2543 | LCC = 0.8538 | SRCC = 0.8536 ]
+INFO:root:Average loss=0.08294013228553992
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2458 | LCC = 0.8495 | SRCC = 0.8504 ]
+INFO:root:Average loss=0.07720066425509942
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2380 | LCC = 0.8510 | SRCC = 0.8518 ]
+INFO:root:Average loss=0.07973640894469543
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2383 | LCC = 0.8531 | SRCC = 0.8532 ]
+INFO:root:Average loss=0.07244081924167964
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2476 | LCC = 0.8460 | SRCC = 0.8452 ]
+INFO:root:Average loss=0.06955408964019555
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2450 | LCC = 0.8496 | SRCC = 0.8494 ]
+INFO:root:Average loss=0.06167036619706032
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2696 | LCC = 0.8440 | SRCC = 0.8436 ]
+INFO:root:Average loss=0.06148498204465096
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2755 | LCC = 0.8439 | SRCC = 0.8430 ]
+INFO:root:Average loss=0.06650885165883945
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2547 | LCC = 0.8477 | SRCC = 0.8471 ]
+INFO:root:Average loss=0.057192399859046325
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2527 | LCC = 0.8450 | SRCC = 0.8445 ]
+INFO:root:Average loss=0.05703940180440744
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2541 | LCC = 0.8491 | SRCC = 0.8501 ]
+INFO:root:Average loss=0.05327783687374531
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2460 | LCC = 0.8481 | SRCC = 0.8479 ]
+INFO:root:Average loss=0.049764264231690995
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2582 | LCC = 0.8453 | SRCC = 0.8447 ]
+INFO:root:Average loss=0.047981671296442166
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2657 | LCC = 0.8437 | SRCC = 0.8419 ]
+INFO:root:Average loss=0.04447451112075494
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2835 | LCC = 0.8452 | SRCC = 0.8450 ]
+INFO:root:Average loss=0.04357468714125645
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2536 | LCC = 0.8481 | SRCC = 0.8473 ]
+INFO:root:Average loss=0.0441230527865581
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2505 | LCC = 0.8434 | SRCC = 0.8419 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2562 | LCC = 0.8669 | SRCC = 0.8678 ]

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/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_8epoch/layer6/run2/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3a340a3c9b00ee154b8cc01d4c70a8607131a1149c8cc7465131a4d8f472467
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/run2/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e949f710450eb03c290163913d30d804fd3d0b850d3fd878d4e4f335753ea46f
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/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=0.35811989085796553
+INFO:root:
+[Valid][0][UTT][ MSE = 0.2855 | LCC = 0.8356 | SRCC = 0.8394 ]
+INFO:root:Average loss=0.23174308660702828
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2696 | LCC = 0.8587 | SRCC = 0.8571 ]
+INFO:root:Average loss=0.20694455829186317
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2411 | LCC = 0.8652 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.19412387449007767
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2507 | LCC = 0.8690 | SRCC = 0.8672 ]
+INFO:root:Average loss=0.17476236848877028
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2181 | LCC = 0.8662 | SRCC = 0.8685 ]
+INFO:root:Average loss=0.16277041458166563
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2422 | LCC = 0.8689 | SRCC = 0.8669 ]
+INFO:root:Average loss=0.14799939946104318
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2368 | LCC = 0.8692 | SRCC = 0.8682 ]
+INFO:root:Average loss=0.14131302386522293
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2461 | LCC = 0.8671 | SRCC = 0.8668 ]
+INFO:root:Average loss=0.12920137915091637
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2390 | LCC = 0.8635 | SRCC = 0.8640 ]
+INFO:root:Average loss=0.12187185606513268
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2271 | LCC = 0.8669 | SRCC = 0.8675 ]
+INFO:root:Average loss=0.1094062059926681
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2259 | LCC = 0.8613 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.10354185056609985
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2506 | LCC = 0.8603 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.09743791331465428
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2289 | LCC = 0.8613 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.09324750246909949
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2974 | LCC = 0.8539 | SRCC = 0.8557 ]
+INFO:root:Average loss=0.08660409962519622
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2674 | LCC = 0.8559 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.0802232102992443
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2495 | LCC = 0.8566 | SRCC = 0.8564 ]
+INFO:root:Average loss=0.07793291405034371
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2467 | LCC = 0.8566 | SRCC = 0.8556 ]
+INFO:root:Average loss=0.07416404583133183
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2503 | LCC = 0.8575 | SRCC = 0.8589 ]
+INFO:root:Average loss=0.07246936599795635
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2352 | LCC = 0.8545 | SRCC = 0.8548 ]
+INFO:root:Average loss=0.06706997064443734
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2268 | LCC = 0.8561 | SRCC = 0.8555 ]
+INFO:root:Average loss=0.06369093447350539
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2389 | LCC = 0.8555 | SRCC = 0.8549 ]
+INFO:root:Average loss=0.057506588072730944
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2496 | LCC = 0.8510 | SRCC = 0.8510 ]
+INFO:root:Average loss=0.06347753059787628
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2309 | LCC = 0.8569 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.055077012532796614
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2539 | LCC = 0.8537 | SRCC = 0.8541 ]
+INFO:root:Average loss=0.05302810387160534
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2626 | LCC = 0.8524 | SRCC = 0.8529 ]
+INFO:root:Average loss=0.05048807672201059
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2428 | LCC = 0.8513 | SRCC = 0.8510 ]
+INFO:root:Average loss=0.05132855040331682
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2463 | LCC = 0.8525 | SRCC = 0.8521 ]
+INFO:root:Average loss=0.046166109470411755
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2623 | LCC = 0.8507 | SRCC = 0.8507 ]
+INFO:root:Average loss=0.04295750248890657
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2365 | LCC = 0.8521 | SRCC = 0.8527 ]
+INFO:root:Average loss=0.04421266321188364
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2431 | LCC = 0.8536 | SRCC = 0.8542 ]
+INFO:root:Average loss=0.04202679036041865
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2473 | LCC = 0.8542 | SRCC = 0.8547 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2523 | LCC = 0.8611 | SRCC = 0.8646 ]

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/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_8epoch/layer6/run3/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90a5e9750e159716c95c09fb2e3ad10009d90142d3ecc0dc5c7137a955376e50
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/run3/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4d858653d04b8c64d1c8ce6ece083d1c3e874c6c955e93f2d0707ce68f7482e
+size 2749638

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/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=0.3604837806943135
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3241 | LCC = 0.8343 | SRCC = 0.8350 ]
+INFO:root:Average loss=0.2331581534101413
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2768 | LCC = 0.8506 | SRCC = 0.8496 ]
+INFO:root:Average loss=0.2057274090938079
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2501 | LCC = 0.8595 | SRCC = 0.8580 ]
+INFO:root:Average loss=0.191218614196166
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2560 | LCC = 0.8631 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.17719015383567566
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2326 | LCC = 0.8634 | SRCC = 0.8630 ]
+INFO:root:Average loss=0.17022792326334196
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2758 | LCC = 0.8614 | SRCC = 0.8604 ]
+INFO:root:Average loss=0.1579711240453598
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2327 | LCC = 0.8606 | SRCC = 0.8610 ]
+INFO:root:Average loss=0.14690540969753876
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2559 | LCC = 0.8593 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.13715836730522987
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2581 | LCC = 0.8587 | SRCC = 0.8578 ]
+INFO:root:Average loss=0.12865965249828804
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2611 | LCC = 0.8561 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.11679353803778306
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2572 | LCC = 0.8575 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.1120085115902699
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2507 | LCC = 0.8551 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.10751390299544884
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2619 | LCC = 0.8516 | SRCC = 0.8521 ]
+INFO:root:Average loss=0.10514366779571925
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2485 | LCC = 0.8506 | SRCC = 0.8511 ]
+INFO:root:Average loss=0.09369600846026188
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2562 | LCC = 0.8512 | SRCC = 0.8518 ]
+INFO:root:Average loss=0.08671288693753573
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2629 | LCC = 0.8497 | SRCC = 0.8518 ]
+INFO:root:Average loss=0.08395195432389393
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2715 | LCC = 0.8498 | SRCC = 0.8515 ]
+INFO:root:Average loss=0.08351552882828774
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2751 | LCC = 0.8487 | SRCC = 0.8502 ]
+INFO:root:Average loss=0.08010196704895069
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2838 | LCC = 0.8459 | SRCC = 0.8481 ]
+INFO:root:Average loss=0.0744340203415889
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2526 | LCC = 0.8501 | SRCC = 0.8512 ]
+INFO:root:Average loss=0.06690540408285764
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2752 | LCC = 0.8450 | SRCC = 0.8464 ]
+INFO:root:Average loss=0.07163178252104001
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2750 | LCC = 0.8497 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.06305020106717563
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2566 | LCC = 0.8493 | SRCC = 0.8493 ]
+INFO:root:Average loss=0.05843058061332275
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2525 | LCC = 0.8448 | SRCC = 0.8452 ]
+INFO:root:Average loss=0.05796974214414755
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2625 | LCC = 0.8451 | SRCC = 0.8460 ]
+INFO:root:Average loss=0.05792016645845694
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2687 | LCC = 0.8429 | SRCC = 0.8440 ]
+INFO:root:Average loss=0.0553492755175401
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2491 | LCC = 0.8500 | SRCC = 0.8518 ]
+INFO:root:Average loss=0.05292717873668059
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2502 | LCC = 0.8471 | SRCC = 0.8483 ]
+INFO:root:Average loss=0.04914245057182434
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2660 | LCC = 0.8468 | SRCC = 0.8488 ]
+INFO:root:Average loss=0.046933587974844836
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2531 | LCC = 0.8474 | SRCC = 0.8484 ]
+INFO:root:Average loss=0.04748835655836723
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2588 | LCC = 0.8448 | SRCC = 0.8453 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2281 | LCC = 0.8657 | SRCC = 0.8675 ]

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/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_8epoch/layer6/run4/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:397f58628d505f682ce8eef89e81c31956362c4e51ff7457d979973ae6441cfe
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/run4/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5aed4c0e3a4ffc26b47e238b8ba399c924795cf8d7087008e04d50b87280e34e
+size 2749638

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer6/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=0.3770978162303949
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3275 | LCC = 0.8323 | SRCC = 0.8332 ]
+INFO:root:Average loss=0.2387025990547278
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2596 | LCC = 0.8497 | SRCC = 0.8498 ]
+INFO:root:Average loss=0.21004770352290228
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2532 | LCC = 0.8611 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.1903307160887963
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2562 | LCC = 0.8654 | SRCC = 0.8620 ]
+INFO:root:Average loss=0.17257102416493955
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2605 | LCC = 0.8643 | SRCC = 0.8634 ]
+INFO:root:Average loss=0.16535658886035284
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2388 | LCC = 0.8655 | SRCC = 0.8653 ]
+INFO:root:Average loss=0.15282277906170258
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2265 | LCC = 0.8653 | SRCC = 0.8646 ]
+INFO:root:Average loss=0.1427481387479183
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2672 | LCC = 0.8649 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.13475456872047523
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2567 | LCC = 0.8611 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.12519202104363686
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2382 | LCC = 0.8615 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.11535744235301629
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2624 | LCC = 0.8616 | SRCC = 0.8614 ]
+INFO:root:Average loss=0.10666486000021298
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2487 | LCC = 0.8593 | SRCC = 0.8570 ]
+INFO:root:Average loss=0.1015381939136065
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2378 | LCC = 0.8577 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.0960148461163044
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2706 | LCC = 0.8567 | SRCC = 0.8539 ]
+INFO:root:Average loss=0.09119731350204884
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2399 | LCC = 0.8560 | SRCC = 0.8547 ]
+INFO:root:Average loss=0.081835809808511
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2524 | LCC = 0.8541 | SRCC = 0.8513 ]
+INFO:root:Average loss=0.08057321904179378
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2529 | LCC = 0.8497 | SRCC = 0.8479 ]
+INFO:root:Average loss=0.07568612451163623
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2479 | LCC = 0.8510 | SRCC = 0.8491 ]
+INFO:root:Average loss=0.070625869748302
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2463 | LCC = 0.8553 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.06824355677534373
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2423 | LCC = 0.8536 | SRCC = 0.8517 ]
+INFO:root:Average loss=0.06662077552232987
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2452 | LCC = 0.8556 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.059250764978619724
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2390 | LCC = 0.8555 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.058074076349536576
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2421 | LCC = 0.8544 | SRCC = 0.8522 ]
+INFO:root:Average loss=0.05762129979064831
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2617 | LCC = 0.8510 | SRCC = 0.8492 ]
+INFO:root:Average loss=0.056809222230162375
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2373 | LCC = 0.8537 | SRCC = 0.8508 ]
+INFO:root:Average loss=0.05110459559811996
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2363 | LCC = 0.8545 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.05058353465910141
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2560 | LCC = 0.8498 | SRCC = 0.8479 ]
+INFO:root:Average loss=0.04859237482723517
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2396 | LCC = 0.8527 | SRCC = 0.8519 ]
+INFO:root:Average loss=0.04709915105157938
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2621 | LCC = 0.8506 | SRCC = 0.8497 ]
+INFO:root:Average loss=0.04966300873993299
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2490 | LCC = 0.8528 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.04379731996987875
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2431 | LCC = 0.8532 | SRCC = 0.8525 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2391 | LCC = 0.8677 | SRCC = 0.8670 ]

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/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_8epoch/layer7/run0/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcb23afd3177b4134fd8992391887bb35558ec5236e983a5ea6ef140639c1514
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/run0/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d95505fe3bca9aafb0219284b64b790e45615f1a520ecad914672b76849e621
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/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=0.3469079965964342
+INFO:root:
+[Valid][0][UTT][ MSE = 0.2989 | LCC = 0.8382 | SRCC = 0.8401 ]
+INFO:root:Average loss=0.23198478783552462
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2585 | LCC = 0.8558 | SRCC = 0.8529 ]
+INFO:root:Average loss=0.20541685781417748
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2483 | LCC = 0.8641 | SRCC = 0.8621 ]
+INFO:root:Average loss=0.19130761921405792
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2713 | LCC = 0.8677 | SRCC = 0.8651 ]
+INFO:root:Average loss=0.18461942663177466
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2428 | LCC = 0.8685 | SRCC = 0.8679 ]
+INFO:root:Average loss=0.16500769097071427
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2516 | LCC = 0.8690 | SRCC = 0.8660 ]
+INFO:root:Average loss=0.15804373329648605
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2455 | LCC = 0.8693 | SRCC = 0.8679 ]
+INFO:root:Average loss=0.15088937603510344
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2357 | LCC = 0.8691 | SRCC = 0.8673 ]
+INFO:root:Average loss=0.1387404889250413
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2327 | LCC = 0.8666 | SRCC = 0.8658 ]
+INFO:root:Average loss=0.13043687473504972
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2570 | LCC = 0.8636 | SRCC = 0.8635 ]
+INFO:root:Average loss=0.11843642649742273
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2293 | LCC = 0.8645 | SRCC = 0.8642 ]
+INFO:root:Average loss=0.12040004864908181
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2339 | LCC = 0.8614 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.11031159605735387
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2628 | LCC = 0.8545 | SRCC = 0.8545 ]
+INFO:root:Average loss=0.10428104143685256
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2402 | LCC = 0.8574 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.09424971884641892
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2438 | LCC = 0.8565 | SRCC = 0.8566 ]
+INFO:root:Average loss=0.08903348732453126
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2415 | LCC = 0.8514 | SRCC = 0.8501 ]
+INFO:root:Average loss=0.08564536870481113
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2446 | LCC = 0.8547 | SRCC = 0.8534 ]
+INFO:root:Average loss=0.08497493747526254
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2353 | LCC = 0.8549 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.07745299349801663
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2783 | LCC = 0.8477 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.07135221438530163
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2526 | LCC = 0.8502 | SRCC = 0.8495 ]
+INFO:root:Average loss=0.07131569780027255
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2605 | LCC = 0.8463 | SRCC = 0.8451 ]
+INFO:root:Average loss=0.06440688287600493
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2764 | LCC = 0.8484 | SRCC = 0.8474 ]
+INFO:root:Average loss=0.0627313334112748
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2873 | LCC = 0.8452 | SRCC = 0.8444 ]
+INFO:root:Average loss=0.05825639129258119
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2651 | LCC = 0.8423 | SRCC = 0.8421 ]
+INFO:root:Average loss=0.056449636578177795
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2629 | LCC = 0.8450 | SRCC = 0.8436 ]
+INFO:root:Average loss=0.055312554948987104
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2510 | LCC = 0.8484 | SRCC = 0.8479 ]
+INFO:root:Average loss=0.051834007199758135
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2352 | LCC = 0.8502 | SRCC = 0.8500 ]
+INFO:root:Average loss=0.04881216726528528
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2513 | LCC = 0.8447 | SRCC = 0.8440 ]
+INFO:root:Average loss=0.04795687951338597
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2557 | LCC = 0.8469 | SRCC = 0.8456 ]
+INFO:root:Average loss=0.045337436434168085
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2537 | LCC = 0.8483 | SRCC = 0.8473 ]
+INFO:root:Average loss=0.04486588393457425
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2551 | LCC = 0.8463 | SRCC = 0.8444 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2494 | LCC = 0.8673 | SRCC = 0.8665 ]

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/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_8epoch/layer7/run1/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53753beeb216822b7a44963f08a62439e29026986a49876121c9e28db9c01ac6
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/run1/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22e32b5b8e6cff4723c3767c557a1cf8849e21f40cde73a0161d2cbd38cc58cc
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/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=0.35092105181553424
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3467 | LCC = 0.8330 | SRCC = 0.8389 ]
+INFO:root:Average loss=0.22391803409808722
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2580 | LCC = 0.8565 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.2077186908095311
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2649 | LCC = 0.8597 | SRCC = 0.8611 ]
+INFO:root:Average loss=0.1873732989606185
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2662 | LCC = 0.8629 | SRCC = 0.8674 ]
+INFO:root:Average loss=0.1828128002010859
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2472 | LCC = 0.8694 | SRCC = 0.8691 ]
+INFO:root:Average loss=0.17032867536330834
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2253 | LCC = 0.8675 | SRCC = 0.8682 ]
+INFO:root:Average loss=0.15470075215666723
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2232 | LCC = 0.8677 | SRCC = 0.8674 ]
+INFO:root:Average loss=0.14759624739870048
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2272 | LCC = 0.8673 | SRCC = 0.8694 ]
+INFO:root:Average loss=0.14010418645846537
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2238 | LCC = 0.8671 | SRCC = 0.8693 ]
+INFO:root:Average loss=0.1272563649675785
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2293 | LCC = 0.8661 | SRCC = 0.8664 ]
+INFO:root:Average loss=0.12291097459502709
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2458 | LCC = 0.8646 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.11045287401439288
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2423 | LCC = 0.8621 | SRCC = 0.8611 ]
+INFO:root:Average loss=0.10909775281563783
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2404 | LCC = 0.8589 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.09805169773216431
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2434 | LCC = 0.8584 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.09028928206326106
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2443 | LCC = 0.8571 | SRCC = 0.8571 ]
+INFO:root:Average loss=0.08622201809134239
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2455 | LCC = 0.8500 | SRCC = 0.8495 ]
+INFO:root:Average loss=0.08073516118411835
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2566 | LCC = 0.8536 | SRCC = 0.8546 ]
+INFO:root:Average loss=0.07807559548662259
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2554 | LCC = 0.8526 | SRCC = 0.8529 ]
+INFO:root:Average loss=0.07593226824433376
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2307 | LCC = 0.8567 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.07104522601152077
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2343 | LCC = 0.8569 | SRCC = 0.8568 ]
+INFO:root:Average loss=0.0637034338254195
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2489 | LCC = 0.8500 | SRCC = 0.8502 ]
+INFO:root:Average loss=0.06301592171001129
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2420 | LCC = 0.8489 | SRCC = 0.8498 ]
+INFO:root:Average loss=0.05648809709610083
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2530 | LCC = 0.8474 | SRCC = 0.8473 ]
+INFO:root:Average loss=0.0569760552010475
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2716 | LCC = 0.8537 | SRCC = 0.8538 ]
+INFO:root:Average loss=0.054729500642189614
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2478 | LCC = 0.8465 | SRCC = 0.8482 ]
+INFO:root:Average loss=0.05050178380826345
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2516 | LCC = 0.8517 | SRCC = 0.8526 ]
+INFO:root:Average loss=0.0497330358872811
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2371 | LCC = 0.8499 | SRCC = 0.8501 ]
+INFO:root:Average loss=0.04755664182205995
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2439 | LCC = 0.8511 | SRCC = 0.8500 ]
+INFO:root:Average loss=0.046851443532758795
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2498 | LCC = 0.8515 | SRCC = 0.8519 ]
+INFO:root:Average loss=0.04367782999403202
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2481 | LCC = 0.8487 | SRCC = 0.8492 ]
+INFO:root:Average loss=0.04227986633300017
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2561 | LCC = 0.8521 | SRCC = 0.8512 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2591 | LCC = 0.8659 | SRCC = 0.8697 ]

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/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_8epoch/layer7/run2/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:808f410352b17bde41dd47121471350a76295d454545e44e38a25e69823b3dff
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/run2/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e50b120f4936b41a4e400e3d81050ab5235850aa90ab0561a6fd79bd6b6b76f9
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/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=0.3361775741363183
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3013 | LCC = 0.8406 | SRCC = 0.8431 ]
+INFO:root:Average loss=0.22454409377697188
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3010 | LCC = 0.8625 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.20117125765253335
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2633 | LCC = 0.8674 | SRCC = 0.8654 ]
+INFO:root:Average loss=0.18968014877576095
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2349 | LCC = 0.8672 | SRCC = 0.8676 ]
+INFO:root:Average loss=0.17348546171799684
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2443 | LCC = 0.8684 | SRCC = 0.8675 ]
+INFO:root:Average loss=0.1605497519366252
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2410 | LCC = 0.8680 | SRCC = 0.8677 ]
+INFO:root:Average loss=0.16413550499158028
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2729 | LCC = 0.8683 | SRCC = 0.8689 ]
+INFO:root:Average loss=0.1432249536498999
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2575 | LCC = 0.8671 | SRCC = 0.8662 ]
+INFO:root:Average loss=0.13707019159427056
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2333 | LCC = 0.8670 | SRCC = 0.8674 ]
+INFO:root:Average loss=0.12744018387718078
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2412 | LCC = 0.8634 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.11622359966620421
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2643 | LCC = 0.8639 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.11292111424681468
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2803 | LCC = 0.8580 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.10526881997401898
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2425 | LCC = 0.8621 | SRCC = 0.8636 ]
+INFO:root:Average loss=0.10095537816866851
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2605 | LCC = 0.8557 | SRCC = 0.8590 ]
+INFO:root:Average loss=0.09372408124498832
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2570 | LCC = 0.8530 | SRCC = 0.8542 ]
+INFO:root:Average loss=0.08658124324985039
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2685 | LCC = 0.8596 | SRCC = 0.8607 ]
+INFO:root:Average loss=0.0811407636755552
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2395 | LCC = 0.8570 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.07503594157214348
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2475 | LCC = 0.8559 | SRCC = 0.8550 ]
+INFO:root:Average loss=0.07276743564468163
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2805 | LCC = 0.8557 | SRCC = 0.8550 ]
+INFO:root:Average loss=0.07363468394256555
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2336 | LCC = 0.8569 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.07126202931006749
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2502 | LCC = 0.8539 | SRCC = 0.8541 ]
+INFO:root:Average loss=0.0614911320213324
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2507 | LCC = 0.8570 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.06123488305661923
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2405 | LCC = 0.8536 | SRCC = 0.8550 ]
+INFO:root:Average loss=0.05606937422775305
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2599 | LCC = 0.8507 | SRCC = 0.8510 ]
+INFO:root:Average loss=0.05564926383204949
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2626 | LCC = 0.8530 | SRCC = 0.8536 ]
+INFO:root:Average loss=0.05060609640219273
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2654 | LCC = 0.8502 | SRCC = 0.8512 ]
+INFO:root:Average loss=0.05456724017858505
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2618 | LCC = 0.8515 | SRCC = 0.8526 ]
+INFO:root:Average loss=0.05297986059807814
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2483 | LCC = 0.8501 | SRCC = 0.8511 ]
+INFO:root:Average loss=0.04690793209159986
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2490 | LCC = 0.8527 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.04672756762458728
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2377 | LCC = 0.8532 | SRCC = 0.8547 ]
+INFO:root:Average loss=0.051660896016237065
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2572 | LCC = 0.8520 | SRCC = 0.8522 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2382 | LCC = 0.8607 | SRCC = 0.8628 ]

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/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_8epoch/layer7/run3/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49619c4ec3b71f6a7b998d5fb817bd1de77e54dd5338a2cba438d42e85a6f417
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/run3/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f6d6aff373ff50a73cacdb16fce4025d6a77eaf15c43ae9bd6556f1aa0d0cc2
+size 2749702

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/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=0.36031487526801914
+INFO:root:
+[Valid][0][UTT][ MSE = 0.2909 | LCC = 0.8361 | SRCC = 0.8392 ]
+INFO:root:Average loss=0.22986766352103308
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2835 | LCC = 0.8539 | SRCC = 0.8506 ]
+INFO:root:Average loss=0.21281702033220193
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2674 | LCC = 0.8612 | SRCC = 0.8583 ]
+INFO:root:Average loss=0.19140471250582963
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2410 | LCC = 0.8623 | SRCC = 0.8600 ]
+INFO:root:Average loss=0.18375769677834633
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2631 | LCC = 0.8599 | SRCC = 0.8586 ]
+INFO:root:Average loss=0.17224495705121604
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2364 | LCC = 0.8620 | SRCC = 0.8609 ]
+INFO:root:Average loss=0.15913706100904024
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2423 | LCC = 0.8638 | SRCC = 0.8621 ]
+INFO:root:Average loss=0.155670099820082
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2358 | LCC = 0.8633 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.14069017166128525
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2496 | LCC = 0.8571 | SRCC = 0.8553 ]
+INFO:root:Average loss=0.1286107660868229
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2474 | LCC = 0.8587 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.12176207672709073
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2495 | LCC = 0.8587 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.11548341027437112
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2463 | LCC = 0.8580 | SRCC = 0.8565 ]
+INFO:root:Average loss=0.10905310291892444
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2554 | LCC = 0.8566 | SRCC = 0.8552 ]
+INFO:root:Average loss=0.10277998423538147
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2511 | LCC = 0.8490 | SRCC = 0.8482 ]
+INFO:root:Average loss=0.09820752328213973
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2500 | LCC = 0.8500 | SRCC = 0.8498 ]
+INFO:root:Average loss=0.09047077357386932
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2468 | LCC = 0.8519 | SRCC = 0.8514 ]
+INFO:root:Average loss=0.08370262503815003
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2386 | LCC = 0.8501 | SRCC = 0.8502 ]
+INFO:root:Average loss=0.0783373729731792
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2434 | LCC = 0.8489 | SRCC = 0.8481 ]
+INFO:root:Average loss=0.07786965699723133
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2333 | LCC = 0.8554 | SRCC = 0.8550 ]
+INFO:root:Average loss=0.07153331612547238
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2397 | LCC = 0.8520 | SRCC = 0.8514 ]
+INFO:root:Average loss=0.06711864270842992
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2385 | LCC = 0.8527 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.06474789690512878
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2565 | LCC = 0.8452 | SRCC = 0.8442 ]
+INFO:root:Average loss=0.06640106401382348
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2527 | LCC = 0.8476 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.0617793684777541
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2416 | LCC = 0.8485 | SRCC = 0.8480 ]
+INFO:root:Average loss=0.05555884659481354
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2528 | LCC = 0.8471 | SRCC = 0.8463 ]
+INFO:root:Average loss=0.05438259654702284
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2425 | LCC = 0.8511 | SRCC = 0.8501 ]
+INFO:root:Average loss=0.050841256307485774
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2370 | LCC = 0.8497 | SRCC = 0.8503 ]
+INFO:root:Average loss=0.05048329660143608
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2482 | LCC = 0.8499 | SRCC = 0.8490 ]
+INFO:root:Average loss=0.047865212871096074
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2448 | LCC = 0.8464 | SRCC = 0.8446 ]
+INFO:root:Average loss=0.04802763521766815
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2605 | LCC = 0.8441 | SRCC = 0.8437 ]
+INFO:root:Average loss=0.0466090687431204
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2672 | LCC = 0.8404 | SRCC = 0.8388 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2572 | LCC = 0.8572 | SRCC = 0.8603 ]

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/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_8epoch/layer7/run4/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b91b223ae33e3bae2145beadebba1e4d53b06115e58c1332019eace278d7cdca
+size 2747979

results_bvcc/w2v2_base_bvcc_finetuned_8epoch/layer7/run4/model_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b64ab0e2aaa9ab1dff702b1235d92f51ae9897c2669acc2fb94579a4617f413
+size 2749702